Machine Learning Approaches to Predict OCD Symptoms
Caitlin Stamatis
5/08/2020
Distinguishing OCD patients vs. healthy controls
Select feature subset & impute missing data
Features chosen based on neuropsych related to INC/HA in context of frontostriatal model (c.f. Bragdon & Coles). Only features with less than 20% missingness were included.
## pobs
## AST.Congruency.cost..Mean..correct. 0.5838926
## AST.Switching.cost..Mean..correct. 0.5973154
## AST.Mean.correct.latency 0.5973154
## AST.Mean.correct.latency..congruent. 0.5973154
## AST.Mean.correct.latency..incongruent. 0.5973154
## AST.Mean.correct.latency..blocks.3.5...non.switching.blocks. 0.5906040
## AST.Mean.correct.latency..block.7...switching.block. 0.5973154
## AST.Percent.correct.trials 0.5838926
## CGT.Delay.aversion 0.7852349
## CGT.Deliberation.time 0.7919463
## CGT.Overall.proportion.bet 0.7986577
## CGT.Quality.of.decision.making 0.7852349
## CGT.Risk.adjustment 0.7919463
## CGT.Risk.taking 0.7986577
## PAL.First.trial.memory.score 0.9731544
## PAL.Stages.completed.on.first.trial 0.9731544
## PAL.Total.errors..adjusted. 0.9194631
## PAL.Mean.errors.to.success 0.9664430
## PAL.Mean.trials.to.success 0.9597315
## PAL.Total.trials..adjusted. 0.9597315
## IED.Total.errors 0.9530201
## IED.Total.errors..adjusted. 0.9597315
## IED.Stages.completed 0.9597315
## IED.Completed.stage.errors 0.9597315
## IED.EDS.errors 0.9664430
## IED.Total.trials..adjusted. 0.9127517
## OTS.Mean.latency.to.correct 0.5234899
## OTS.Mean.latency.to.first.choice 0.5167785
## OTS.Mean.choices.to.correct 0.5302013
## OTS.Problems.solved.on.first.choice 0.5771812
## SRM.Mean.correct.latency 0.7382550
## SRM.Number.correct 0.7449664
## SRM.Percent.correct 0.7449664
## SSP.Number.of.attempts 0.7516779
## SSP.Span.length 0.7516779
## SSP.Mean.time.to.first.response 0.7315436
## SSP.Total.errors 0.7315436
## SST.SSRT..last.half. 0.9664430
## SST.Mean.correct.RT.on.GO.trials 0.9597315
## SST.Direction.errors.on.stop.and.go.trials 0.9463087
## SST.Proportion.of.successful.stops..last.half. 0.9664430
## SST.SSD..50....last.half. 0.9664430
## RVP.A. 0.5906040
## RVP.B.. 0.5906040
## RVP.Probability.of.hit 0.5973154
## RVP.Mean.latency 0.5771812
## RVP.Probability.of.false.alarm 0.5771812
## Pt 1.0000000
## influx
## AST.Congruency.cost..Mean..correct. 0.282478710
## AST.Switching.cost..Mean..correct. 0.265446639
## AST.Mean.correct.latency 0.265446639
## AST.Mean.correct.latency..congruent. 0.265446639
## AST.Mean.correct.latency..incongruent. 0.265446639
## AST.Mean.correct.latency..blocks.3.5...non.switching.blocks. 0.273237905
## AST.Mean.correct.latency..block.7...switching.block. 0.265446639
## AST.Percent.correct.trials 0.281391556
## CGT.Delay.aversion 0.137887298
## CGT.Deliberation.time 0.127015764
## CGT.Overall.proportion.bet 0.122485958
## CGT.Quality.of.decision.making 0.138612067
## CGT.Risk.adjustment 0.130096032
## CGT.Risk.taking 0.122485958
## PAL.First.trial.memory.score 0.008153651
## PAL.Stages.completed.on.first.trial 0.008153651
## PAL.Total.errors..adjusted. 0.063054901
## PAL.Mean.errors.to.success 0.014676572
## PAL.Mean.trials.to.success 0.020112339
## PAL.Total.trials..adjusted. 0.020112339
## IED.Total.errors 0.029534336
## IED.Total.errors..adjusted. 0.025366914
## IED.Stages.completed 0.025366914
## IED.Completed.stage.errors 0.025366914
## IED.EDS.errors 0.018481609
## IED.Total.trials..adjusted. 0.064685631
## OTS.Mean.latency.to.correct 0.338285921
## OTS.Mean.latency.to.first.choice 0.346801957
## OTS.Mean.choices.to.correct 0.331038232
## OTS.Problems.solved.on.first.choice 0.277948904
## SRM.Mean.correct.latency 0.143504258
## SRM.Number.correct 0.134988223
## SRM.Percent.correct 0.134988223
## SSP.Number.of.attempts 0.131726762
## SSP.Span.length 0.131726762
## SSP.Mean.time.to.first.response 0.153832216
## SSP.Total.errors 0.154375793
## SST.SSRT..last.half. 0.018481609
## SST.Mean.correct.RT.on.GO.trials 0.027903606
## SST.Direction.errors.on.stop.and.go.trials 0.041130640
## SST.Proportion.of.successful.stops..last.half. 0.018481609
## SST.SSD..50....last.half. 0.018481609
## RVP.A. 0.271425983
## RVP.B.. 0.271425983
## RVP.Probability.of.hit 0.263272332
## RVP.Mean.latency 0.285921363
## RVP.Probability.of.false.alarm 0.286827324
## Pt 0.000000000
## outflux
## AST.Congruency.cost..Mean..correct. 0.13227189
## AST.Switching.cost..Mean..correct. 0.13349663
## AST.Mean.correct.latency 0.13349663
## AST.Mean.correct.latency..congruent. 0.13349663
## AST.Mean.correct.latency..incongruent. 0.13349663
## AST.Mean.correct.latency..blocks.3.5...non.switching.blocks. 0.13043478
## AST.Mean.correct.latency..block.7...switching.block. 0.13349663
## AST.Percent.correct.trials 0.12859767
## CGT.Delay.aversion 0.52541335
## CGT.Deliberation.time 0.51806491
## CGT.Overall.proportion.bet 0.53214942
## CGT.Quality.of.decision.making 0.52786283
## CGT.Risk.adjustment 0.52847520
## CGT.Risk.taking 0.53214942
## PAL.First.trial.memory.score 0.90998163
## PAL.Stages.completed.on.first.trial 0.90998163
## PAL.Total.errors..adjusted. 0.86037967
## PAL.Mean.errors.to.success 0.90263319
## PAL.Mean.trials.to.success 0.89161053
## PAL.Total.trials..adjusted. 0.89161053
## IED.Total.errors 0.89406001
## IED.Total.errors..adjusted. 0.90936926
## IED.Stages.completed 0.90936926
## IED.Completed.stage.errors 0.90936926
## IED.EDS.errors 0.91549296
## IED.Total.trials..adjusted. 0.83649724
## OTS.Mean.latency.to.correct 0.05633803
## OTS.Mean.latency.to.first.choice 0.05572566
## OTS.Mean.choices.to.correct 0.06123699
## OTS.Problems.solved.on.first.choice 0.08756889
## SRM.Mean.correct.latency 0.33864054
## SRM.Number.correct 0.33925291
## SRM.Percent.correct 0.33925291
## SSP.Number.of.attempts 0.35762400
## SSP.Span.length 0.35762400
## SSP.Mean.time.to.first.response 0.34415187
## SSP.Total.errors 0.34598898
## SST.SSRT..last.half. 0.91549296
## SST.Mean.correct.RT.on.GO.trials 0.91794244
## SST.Direction.errors.on.stop.and.go.trials 0.90385793
## SST.Proportion.of.successful.stops..last.half. 0.91549296
## SST.SSD..50....last.half. 0.91549296
## RVP.A. 0.12431108
## RVP.B.. 0.12431108
## RVP.Probability.of.hit 0.12614819
## RVP.Mean.latency 0.11451317
## RVP.Probability.of.false.alarm 0.11757502
## Pt 1.00000000
## ainb
## AST.Congruency.cost..Mean..correct. 0.5350034
## AST.Switching.cost..Mean..correct. 0.5195035
## AST.Mean.correct.latency 0.5195035
## AST.Mean.correct.latency..congruent. 0.5195035
## AST.Mean.correct.latency..incongruent. 0.5195035
## AST.Mean.correct.latency..blocks.3.5...non.switching.blocks. 0.5259854
## AST.Mean.correct.latency..block.7...switching.block. 0.5195035
## AST.Percent.correct.trials 0.5329444
## CGT.Delay.aversion 0.5059840
## CGT.Deliberation.time 0.4811256
## CGT.Overall.proportion.bet 0.4794326
## CGT.Quality.of.decision.making 0.5086436
## CGT.Risk.adjustment 0.4927934
## CGT.Risk.taking 0.4794326
## PAL.First.trial.memory.score 0.2393617
## PAL.Stages.completed.on.first.trial 0.2393617
## PAL.Total.errors..adjusted. 0.6170213
## PAL.Mean.errors.to.success 0.3446809
## PAL.Mean.trials.to.success 0.3936170
## PAL.Total.trials..adjusted. 0.3936170
## IED.Total.errors 0.4954407
## IED.Total.errors..adjusted. 0.4964539
## IED.Stages.completed 0.4964539
## IED.Completed.stage.errors 0.4964539
## IED.EDS.errors 0.4340426
## IED.Total.trials..adjusted. 0.5842881
## OTS.Mean.latency.to.correct 0.5594846
## OTS.Mean.latency.to.first.choice 0.5656028
## OTS.Mean.choices.to.correct 0.5553191
## OTS.Problems.solved.on.first.choice 0.5180682
## SRM.Mean.correct.latency 0.4320786
## SRM.Number.correct 0.4171333
## SRM.Percent.correct 0.4171333
## SSP.Number.of.attempts 0.4180564
## SSP.Span.length 0.4180564
## SSP.Mean.time.to.first.response 0.4515957
## SSP.Total.errors 0.4531915
## SST.SSRT..last.half. 0.4340426
## SST.Mean.correct.RT.on.GO.trials 0.5460993
## SST.Direction.errors.on.stop.and.go.trials 0.6037234
## SST.Proportion.of.successful.stops..last.half. 0.4340426
## SST.SSD..50....last.half. 0.4340426
## RVP.A. 0.5224974
## RVP.B.. 0.5224974
## RVP.Probability.of.hit 0.5152482
## RVP.Mean.latency 0.5329281
## RVP.Probability.of.false.alarm 0.5346167
## Pt 0.0000000
## aout
## AST.Congruency.cost..Mean..correct. 0.05282465
## AST.Switching.cost..Mean..correct. 0.05211571
## AST.Mean.correct.latency 0.05211571
## AST.Mean.correct.latency..congruent. 0.05211571
## AST.Mean.correct.latency..incongruent. 0.05211571
## AST.Mean.correct.latency..blocks.3.5...non.switching.blocks. 0.05149903
## AST.Mean.correct.latency..block.7...switching.block. 0.05211571
## AST.Percent.correct.trials 0.05135730
## CGT.Delay.aversion 0.15602837
## CGT.Deliberation.time 0.15254237
## CGT.Overall.proportion.bet 0.15537279
## CGT.Quality.of.decision.making 0.15675577
## CGT.Risk.adjustment 0.15560765
## CGT.Risk.taking 0.15537279
## PAL.First.trial.memory.score 0.21804842
## PAL.Stages.completed.on.first.trial 0.21804842
## PAL.Total.errors..adjusted. 0.21820158
## PAL.Mean.errors.to.success 0.21778960
## PAL.Mean.trials.to.success 0.21663443
## PAL.Total.trials..adjusted. 0.21663443
## IED.Total.errors 0.21875936
## IED.Total.errors..adjusted. 0.22094926
## IED.Stages.completed 0.22094926
## IED.Completed.stage.errors 0.22094926
## IED.EDS.errors 0.22089243
## IED.Total.trials..adjusted. 0.21370463
## OTS.Mean.latency.to.correct 0.02509547
## OTS.Mean.latency.to.first.choice 0.02514507
## OTS.Mean.choices.to.correct 0.02693240
## OTS.Problems.solved.on.first.choice 0.03537853
## SRM.Mean.correct.latency 0.10696325
## SRM.Number.correct 0.10619130
## SRM.Percent.correct 0.10619130
## SSP.Number.of.attempts 0.11094225
## SSP.Span.length 0.11094225
## SSP.Mean.time.to.first.response 0.10970135
## SSP.Total.errors 0.11028694
## SST.SSRT..last.half. 0.22089243
## SST.Mean.correct.RT.on.GO.trials 0.22303229
## SST.Direction.errors.on.stop.and.go.trials 0.22272522
## SST.Proportion.of.successful.stops..last.half. 0.22089243
## SST.SSD..50....last.half. 0.22089243
## RVP.A. 0.04908124
## RVP.B.. 0.04908124
## RVP.Probability.of.hit 0.04924695
## RVP.Mean.latency 0.04626423
## RVP.Probability.of.false.alarm 0.04750124
## Pt 0.23318578
## fico
## AST.Congruency.cost..Mean..correct. 0.4252874
## AST.Switching.cost..Mean..correct. 0.4382022
## AST.Mean.correct.latency 0.4382022
## AST.Mean.correct.latency..congruent. 0.4382022
## AST.Mean.correct.latency..incongruent. 0.4382022
## AST.Mean.correct.latency..blocks.3.5...non.switching.blocks. 0.4318182
## AST.Mean.correct.latency..block.7...switching.block. 0.4382022
## AST.Percent.correct.trials 0.4252874
## CGT.Delay.aversion 0.5726496
## CGT.Deliberation.time 0.5762712
## CGT.Overall.proportion.bet 0.5798319
## CGT.Quality.of.decision.making 0.5726496
## CGT.Risk.adjustment 0.5762712
## CGT.Risk.taking 0.5798319
## PAL.First.trial.memory.score 0.6551724
## PAL.Stages.completed.on.first.trial 0.6551724
## PAL.Total.errors..adjusted. 0.6350365
## PAL.Mean.errors.to.success 0.6527778
## PAL.Mean.trials.to.success 0.6503497
## PAL.Total.trials..adjusted. 0.6503497
## IED.Total.errors 0.6478873
## IED.Total.errors..adjusted. 0.6503497
## IED.Stages.completed 0.6503497
## IED.Completed.stage.errors 0.6503497
## IED.EDS.errors 0.6527778
## IED.Total.trials..adjusted. 0.6323529
## OTS.Mean.latency.to.correct 0.3589744
## OTS.Mean.latency.to.first.choice 0.3506494
## OTS.Mean.choices.to.correct 0.3670886
## OTS.Problems.solved.on.first.choice 0.4186047
## SRM.Mean.correct.latency 0.5454545
## SRM.Number.correct 0.5495495
## SRM.Percent.correct 0.5495495
## SSP.Number.of.attempts 0.5535714
## SSP.Span.length 0.5535714
## SSP.Mean.time.to.first.response 0.5412844
## SSP.Total.errors 0.5412844
## SST.SSRT..last.half. 0.6527778
## SST.Mean.correct.RT.on.GO.trials 0.6503497
## SST.Direction.errors.on.stop.and.go.trials 0.6453901
## SST.Proportion.of.successful.stops..last.half. 0.6527778
## SST.SSD..50....last.half. 0.6527778
## RVP.A. 0.4318182
## RVP.B.. 0.4318182
## RVP.Probability.of.hit 0.4382022
## RVP.Mean.latency 0.4186047
## RVP.Probability.of.false.alarm 0.4186047
## Pt 0.6644295
##
## iter imp variable
## 1 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Total.errors..adjusted. PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Stages.completed IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.##
## iter imp variable
## 1 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 1 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 2 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 3 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 4 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 1 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 2 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 3 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 4 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.
## 5 5 CGT.Delay.aversion CGT.Deliberation.time CGT.Overall.proportion.bet CGT.Quality.of.decision.making CGT.Risk.adjustment CGT.Risk.taking PAL.First.trial.memory.score PAL.Stages.completed.on.first.trial PAL.Mean.errors.to.success PAL.Mean.trials.to.success PAL.Total.trials..adjusted. IED.Total.errors IED.Total.errors..adjusted. IED.Completed.stage.errors IED.EDS.errors IED.Total.trials..adjusted. SST.SSRT..last.half. SST.Mean.correct.RT.on.GO.trials SST.Direction.errors.on.stop.and.go.trials SST.Proportion.of.successful.stops..last.half. SST.SSD..50....last.half.Plot distributions of original and imputed data
temat.cog.fs.flux.2%>%
keep(is.numeric) %>% # Keep only numeric columns
gather() %>% # Convert to key-value pairs
ggplot(aes(value)) + # Plot the values
facet_wrap(~ key, scales = "free") + # In separate panels
geom_density() 
temat.cog.fs.flux.2.imputed %>%
keep(is.numeric) %>% # Keep only numeric columns
gather() %>% # Convert to key-value pairs
ggplot(aes(value)) + # Plot the values
facet_wrap(~ key, scales = "free") + # In separate panels
geom_density() 
Random Forest
set.seed(321)
split = sample.split(temat.cog.fs.flux.2.imputed$Pt, SplitRatio = 0.70)
TrainSet = subset(temat.cog.fs.flux.2.imputed, split == TRUE)
ValidSet = subset(temat.cog.fs.flux.2.imputed, split == FALSE)
TrainSet$Pt <- as.factor(as.character(TrainSet$Pt))
ValidSet$Pt <- as.factor(as.character(ValidSet$Pt))
#verify the distribution of 0s/1s in the training/validation sets
prop.table(table(TrainSet$Pt)) * 100##
## 0 1
## 39.42308 60.57692prop.table(table(ValidSet$Pt)) * 100##
## 0 1
## 40 60# 3-fold CV with caret
control <- trainControl(method="cv", number=3, search="random")
# Tune to determine optimal ntree
set.seed(321)
tunegrid <- expand.grid(.mtry=c(sqrt(ncol(TrainSet))))
modellist <- list()
for (ntree in c(51,101,501,601,701,801,901,1001)) {
set.seed(321)
fit <- train(Pt~., data=TrainSet, method="rf", tuneGrid=tunegrid, trControl=control, tunelength=16, ntree=ntree)
key <- toString(ntree)
modellist[[key]] <- fit
}
modellist # highest accuracy: 701 trees## $`51`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6635854 0.2872582
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`101`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6439776 0.2409437
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`501`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6633053 0.2821841
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`601`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6823529 0.3195772
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`701`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6823529 0.3195772
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`801`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6728291 0.300644
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`901`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6728291 0.300644
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`1001`
## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 70, 69, 69
## Resampling results:
##
## Accuracy Kappa
## 0.6728291 0.300644
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416# figure out optimal mtry
set.seed(5)
rf_random <- train(Pt~., data=TrainSet, method="rf", tuneLength=7, trControl=control, ntree=701)
#accuracy in training set
print(rf_random)## Random Forest
##
## 104 samples
## 21 predictor
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 69, 70, 69
## Resampling results across tuning parameters:
##
## mtry Accuracy Kappa
## 3 0.6061625 0.1260846
## 7 0.6061625 0.1496509
## 9 0.6064426 0.1513282
## 11 0.6263305 0.1960337
## 15 0.6257703 0.1999161
## 19 0.6352941 0.2164774
## 21 0.6252101 0.1826977
##
## Accuracy was used to select the optimal model using the largest value.
## The final value used for the model was mtry = 19.confusionMatrix(rf_random$finalModel$predicted, TrainSet$Pt)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 21 14
## 1 20 49
##
## Accuracy : 0.6731
## 95% CI : (0.5741, 0.7619)
## No Information Rate : 0.6058
## P-Value [Acc > NIR] : 0.09506
##
## Kappa : 0.2976
##
## Mcnemar's Test P-Value : 0.39117
##
## Sensitivity : 0.5122
## Specificity : 0.7778
## Pos Pred Value : 0.6000
## Neg Pred Value : 0.7101
## Prevalence : 0.3942
## Detection Rate : 0.2019
## Detection Prevalence : 0.3365
## Balanced Accuracy : 0.6450
##
## 'Positive' Class : 0
## #accuracy in test set
pred_1 = predict(rf_random, ValidSet, type="raw")
confusionMatrix(pred_1, ValidSet$Pt)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 9 6
## 1 9 21
##
## Accuracy : 0.6667
## 95% CI : (0.5105, 0.8)
## No Information Rate : 0.6
## P-Value [Acc > NIR] : 0.2249
##
## Kappa : 0.2857
##
## Mcnemar's Test P-Value : 0.6056
##
## Sensitivity : 0.5000
## Specificity : 0.7778
## Pos Pred Value : 0.6000
## Neg Pred Value : 0.7000
## Prevalence : 0.4000
## Detection Rate : 0.2000
## Detection Prevalence : 0.3333
## Balanced Accuracy : 0.6389
##
## 'Positive' Class : 0
## #variable importance
vi.rf.temat <- varImp(rf_random, scale = T, useModel = TRUE)
plot(vi.rf.temat)
Support Vector Machine
#splitting the data set into the Training set and Test set
set.seed(321)
split = sample.split(temat.cog.fs.flux.2.imputed$Pt, SplitRatio = 0.70)
training_set = subset(temat.cog.fs.flux.2.imputed, split == TRUE)
test_set = subset(temat.cog.fs.flux.2.imputed, split == FALSE)
training_set$Pt <- as.factor(training_set$Pt)
test_set$Pt <- as.factor(test_set$Pt)
#feature scaling
training_set[-22] = scale(training_set[-22])
test_set[-22] = scale(test_set[-22])
prop.table(table(training_set$Pt)) * 100##
## 0 1
## 39.42308 60.57692prop.table(table(test_set$Pt)) * 100##
## 0 1
## 40 60#find optimal tuning parameters for svm assuming radial kernel
trainacc = c() # empty vector to store accuracy for training set
testacc = c() # empty vector to store accuracy for test set
#caret is limited to tuning one kernel at a time--we manually loop through
kernel_list <- list("linear", "polynomial", "radial", "sigmoid")
cost_list <- list(.5,1,1.25,1.4,1.5,1.75,2) # after zeroing in on optimal range
gamma_list <- list(0.125,0.25,0.5,1,2,4,8,16)
n <- 1
for (k in seq_along(kernel_list)){
for (j in seq_along(cost_list)){
for (i in seq_along(gamma_list)){
set.seed(2)
#run model with selected tuning parameters
svm_model_cs_tuned <- svm(Pt~ ., data=training_set, method="C-classification", kernel=kernel_list[[k]],cost=cost_list[[j]],gamma=gamma_list[[i]])
#training set predictions
pred_train <-predict(svm_model_cs_tuned,training_set)
mean(pred_train==training_set$Pt)
trainacc[n] <- mean(pred_train==training_set$Pt)
#test set predictions
pred_test <-predict(svm_model_cs_tuned,test_set)
testacc[n] <- mean(pred_test==test_set$Pt)
n <- n+1
}
}
}
#examine train and test accuracy side by side
kernels <- c(rep("linear",56),rep("polynomial",56),rep("radial",56),rep("sigmoid",56))
costprep <- c(rep(".5",8), rep("1",8),rep("1.25",8),rep("1.4",8),rep("1.5",8),rep("1.75",8),rep("2",8))
costs <- rep(costprep,4)
gammasprep <- c(0.125,0.25,0.5,1,2,4,8,16)
gammas <- c(rep(gammasprep, 28))
accuracytable <- cbind(trainacc, testacc, kernels, gammas, costs)
#accuracytable
# run model using best cost, gamma, kernel and cross-validation
set.seed(2)
svm_model_cs_tuned <- svm(Pt~ ., data=training_set, method="C-classification", kernel="linear",cost=1.5, cross=3)
#training set predictions
pred_train <-predict(svm_model_cs_tuned,training_set)
mean(pred_train==training_set$Pt)## [1] 0.7403846confusionMatrix(pred_train,training_set$Pt)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 23 9
## 1 18 54
##
## Accuracy : 0.7404
## 95% CI : (0.6452, 0.8214)
## No Information Rate : 0.6058
## P-Value [Acc > NIR] : 0.002783
##
## Kappa : 0.4348
##
## Mcnemar's Test P-Value : 0.123658
##
## Sensitivity : 0.5610
## Specificity : 0.8571
## Pos Pred Value : 0.7187
## Neg Pred Value : 0.7500
## Prevalence : 0.3942
## Detection Rate : 0.2212
## Detection Prevalence : 0.3077
## Balanced Accuracy : 0.7091
##
## 'Positive' Class : 0
## #test set predictions
pred_test <-predict(svm_model_cs_tuned,test_set)
mean(pred_test==test_set$Pt)## [1] 0.6confusionMatrix(pred_test,test_set$Pt)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 6 6
## 1 12 21
##
## Accuracy : 0.6
## 95% CI : (0.4433, 0.743)
## No Information Rate : 0.6
## P-Value [Acc > NIR] : 0.5643
##
## Kappa : 0.1176
##
## Mcnemar's Test P-Value : 0.2386
##
## Sensitivity : 0.3333
## Specificity : 0.7778
## Pos Pred Value : 0.5000
## Neg Pred Value : 0.6364
## Prevalence : 0.4000
## Detection Rate : 0.1333
## Detection Prevalence : 0.2667
## Balanced Accuracy : 0.5556
##
## 'Positive' Class : 0
## K Nearest Neighbors
#model tuning
set.seed(321)
ctrl <- trainControl(method="cv",number = 3)
knnFit <- train(Pt ~ ., data = training_set, method = "knn", trControl = ctrl, tuneLength = 2) #optimal k is 5
#training set predictions
pred_train_knn <-predict(knnFit,training_set)
mean(pred_train_knn==training_set$Pt)## [1] 0.7692308confusionMatrix(pred_train_knn,training_set$Pt)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 29 12
## 1 12 51
##
## Accuracy : 0.7692
## 95% CI : (0.6764, 0.8462)
## No Information Rate : 0.6058
## P-Value [Acc > NIR] : 0.0003148
##
## Kappa : 0.5168
##
## Mcnemar's Test P-Value : 1.0000000
##
## Sensitivity : 0.7073
## Specificity : 0.8095
## Pos Pred Value : 0.7073
## Neg Pred Value : 0.8095
## Prevalence : 0.3942
## Detection Rate : 0.2788
## Detection Prevalence : 0.3942
## Balanced Accuracy : 0.7584
##
## 'Positive' Class : 0
## #test set predictions
pred_test_knn <-predict(knnFit,test_set)
mean(pred_test_knn==test_set$Pt)## [1] 0.6confusionMatrix(pred_test_knn,test_set$Pt)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 8 8
## 1 10 19
##
## Accuracy : 0.6
## 95% CI : (0.4433, 0.743)
## No Information Rate : 0.6
## P-Value [Acc > NIR] : 0.5643
##
## Kappa : 0.1509
##
## Mcnemar's Test P-Value : 0.8137
##
## Sensitivity : 0.4444
## Specificity : 0.7037
## Pos Pred Value : 0.5000
## Neg Pred Value : 0.6552
## Prevalence : 0.4000
## Detection Rate : 0.1778
## Detection Prevalence : 0.3556
## Balanced Accuracy : 0.5741
##
## 'Positive' Class : 0
## Testing predictability of OCD patients with and without sensory phenomena
Random Forest
set.seed(321)
split = sample.split(temat.cog.fs.flux.2.imputed$SPSyn, SplitRatio = 0.70)
TrainSet = subset(temat.cog.fs.flux.2.imputed, split == TRUE)
ValidSet = subset(temat.cog.fs.flux.2.imputed, split == FALSE)
TrainSet$SPSyn <- as.factor(as.character(TrainSet$SPSyn))
ValidSet$SPSyn <- as.factor(as.character(ValidSet$SPSyn))
#verify the distribution of 0s/1s in the training/validation sets
prop.table(table(TrainSet$SPSyn)) * 100##
## 0 1
## 61.53846 38.46154prop.table(table(ValidSet$SPSyn)) * 100##
## 0 1
## 60.86957 39.13043# 3-fold CV with caret
control <- trainControl(method="cv", number=3, search="random")
# Tune to determine optimal ntree
set.seed(818)
tunegrid <- expand.grid(.mtry=c(sqrt(ncol(TrainSet))))
modellist <- list()
for (ntree in c(51,101,501,601,701,801,901,1001,1101)) {
set.seed(123)
fit <- train(SPSyn~., data=TrainSet, method="rf", tuneGrid=tunegrid, trControl=control, tunelength=8, ntree=ntree)
key <- toString(ntree)
modellist[[key]] <- fit
}
modellist # highest accuracy: 601 trees## $`51`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.5648148 0.02497972
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`101`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.5462963 -0.005697499
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`501`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.6018519 0.1029018
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`601`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.6018519 0.1029018
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`701`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.5810185 0.04136333
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`801`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.6018519 0.1029018
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`901`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.6018519 0.1029018
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`1001`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.6018519 0.1029018
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416
##
## $`1101`
## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results:
##
## Accuracy Kappa
## 0.5833333 0.0545901
##
## Tuning parameter 'mtry' was held constant at a value of 4.690416set.seed(123)
rf_random <- train(SPSyn~., data=TrainSet, method="rf", tuneLength=8, trControl=control, ntree=601)
print(rf_random)## Random Forest
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 36, 34
## Resampling results across tuning parameters:
##
## mtry Accuracy Kappa
## 3 0.5625000 -0.006948357
## 5 0.5810185 0.044566210
## 10 0.6203704 0.154416360
## 11 0.6018519 0.123183483
## 14 0.6018519 0.123183483
## 18 0.6203704 0.154871795
## 19 0.6203704 0.171495171
## 20 0.6018519 0.103912891
##
## Accuracy was used to select the optimal model using the largest value.
## The final value used for the model was mtry = 10.confusionMatrix(rf_random$finalModel$predicted, TrainSet$SPSyn)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 23 17
## 1 9 3
##
## Accuracy : 0.5
## 95% CI : (0.3581, 0.6419)
## No Information Rate : 0.6154
## P-Value [Acc > NIR] : 0.9667
##
## Kappa : -0.1419
##
## Mcnemar's Test P-Value : 0.1698
##
## Sensitivity : 0.7188
## Specificity : 0.1500
## Pos Pred Value : 0.5750
## Neg Pred Value : 0.2500
## Prevalence : 0.6154
## Detection Rate : 0.4423
## Detection Prevalence : 0.7692
## Balanced Accuracy : 0.4344
##
## 'Positive' Class : 0
## pred_1 = predict(rf_random, ValidSet, type="raw")
confusionMatrix(pred_1, ValidSet$SPSyn)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 11 8
## 1 3 1
##
## Accuracy : 0.5217
## 95% CI : (0.3059, 0.7318)
## No Information Rate : 0.6087
## P-Value [Acc > NIR] : 0.8570
##
## Kappa : -0.1145
##
## Mcnemar's Test P-Value : 0.2278
##
## Sensitivity : 0.7857
## Specificity : 0.1111
## Pos Pred Value : 0.5789
## Neg Pred Value : 0.2500
## Prevalence : 0.6087
## Detection Rate : 0.4783
## Detection Prevalence : 0.8261
## Balanced Accuracy : 0.4484
##
## 'Positive' Class : 0
## Support Vector Machine
#splitting the data set into the Training set and Test set
set.seed(321)
split = sample.split(temat.cog.fs.flux.2.imputed$SPSyn, SplitRatio = 0.70)
training_set = subset(temat.cog.fs.flux.2.imputed, split == TRUE)
test_set = subset(temat.cog.fs.flux.2.imputed, split == FALSE)
training_set$SPSyn <- as.factor(training_set$SPSyn)
test_set$SPSyn <- as.factor(test_set$SPSyn)
#feature scaling
training_set[-22] = scale(training_set[-22])
test_set[-22] = scale(test_set[-22])
prop.table(table(training_set$SPSyn)) * 100##
## 0 1
## 61.53846 38.46154prop.table(table(test_set$SPSyn)) * 100##
## 0 1
## 60.86957 39.13043#find optimal tuning parameters for svm assuming radial kernel
#Caret limited to tuning 1 parameter at a time - manually loop through
trainacc = c()
testacc = c()
kernel_list <- list("linear", "polynomial", "radial", "sigmoid")
cost_list <- list(.5,1,1.5, 2, 2.5,3,3.5) # these values are after having zeroed in on optimal cost parameter range
gamma_list <- list(0.125,0.25,0.5,1,2,4,8,16)
n <- 1
for (k in seq_along(kernel_list)){
for (j in seq_along(cost_list)){
for (i in seq_along(gamma_list)){
set.seed(321)
#run model with selected tuning parameters
svm_model_cs_tuned <- svm(SPSyn~ ., data=training_set, method="C-classification", kernel=kernel_list[[k]],cost=cost_list[[j]],gamma=gamma_list[[i]])
#training set predictions
pred_train <-predict(svm_model_cs_tuned,training_set)
mean(pred_train==training_set$SPSyn)
trainacc[n] <- mean(pred_train==training_set$SPSyn)
#test set predictions
pred_test <-predict(svm_model_cs_tuned,test_set)
testacc[n] <- mean(pred_test==test_set$SPSyn)
n <- n+1
}
}
}
#examine train and test accuracy side by side
kernels <- c(rep("linear",56),rep("polynomial",56),rep("radial",56),rep("sigmoid",56))
costprep <- c(rep(".5",8), rep("1",8),rep("1.5",8),rep("2",8),rep("2.5",8),rep("3",8),rep("3.5",8))
costs <- rep(costprep,4)
gammasprep <- c(0.125,0.25,0.5,1,2,4,8,16)
gammas <- c(rep(gammasprep, 28))
accuracytable <- cbind(trainacc, testacc, kernels, gammas, costs)
accuracytable## trainacc testacc kernels gammas costs
## [1,] "0.788461538461538" "0.347826086956522" "linear" "0.125" ".5"
## [2,] "0.788461538461538" "0.347826086956522" "linear" "0.25" ".5"
## [3,] "0.788461538461538" "0.347826086956522" "linear" "0.5" ".5"
## [4,] "0.788461538461538" "0.347826086956522" "linear" "1" ".5"
## [5,] "0.788461538461538" "0.347826086956522" "linear" "2" ".5"
## [6,] "0.788461538461538" "0.347826086956522" "linear" "4" ".5"
## [7,] "0.788461538461538" "0.347826086956522" "linear" "8" ".5"
## [8,] "0.788461538461538" "0.347826086956522" "linear" "16" ".5"
## [9,] "0.846153846153846" "0.347826086956522" "linear" "0.125" "1"
## [10,] "0.846153846153846" "0.347826086956522" "linear" "0.25" "1"
## [11,] "0.846153846153846" "0.347826086956522" "linear" "0.5" "1"
## [12,] "0.846153846153846" "0.347826086956522" "linear" "1" "1"
## [13,] "0.846153846153846" "0.347826086956522" "linear" "2" "1"
## [14,] "0.846153846153846" "0.347826086956522" "linear" "4" "1"
## [15,] "0.846153846153846" "0.347826086956522" "linear" "8" "1"
## [16,] "0.846153846153846" "0.347826086956522" "linear" "16" "1"
## [17,] "0.865384615384615" "0.304347826086957" "linear" "0.125" "1.5"
## [18,] "0.865384615384615" "0.304347826086957" "linear" "0.25" "1.5"
## [19,] "0.865384615384615" "0.304347826086957" "linear" "0.5" "1.5"
## [20,] "0.865384615384615" "0.304347826086957" "linear" "1" "1.5"
## [21,] "0.865384615384615" "0.304347826086957" "linear" "2" "1.5"
## [22,] "0.865384615384615" "0.304347826086957" "linear" "4" "1.5"
## [23,] "0.865384615384615" "0.304347826086957" "linear" "8" "1.5"
## [24,] "0.865384615384615" "0.304347826086957" "linear" "16" "1.5"
## [25,] "0.865384615384615" "0.347826086956522" "linear" "0.125" "2"
## [26,] "0.865384615384615" "0.347826086956522" "linear" "0.25" "2"
## [27,] "0.865384615384615" "0.347826086956522" "linear" "0.5" "2"
## [28,] "0.865384615384615" "0.347826086956522" "linear" "1" "2"
## [29,] "0.865384615384615" "0.347826086956522" "linear" "2" "2"
## [30,] "0.865384615384615" "0.347826086956522" "linear" "4" "2"
## [31,] "0.865384615384615" "0.347826086956522" "linear" "8" "2"
## [32,] "0.865384615384615" "0.347826086956522" "linear" "16" "2"
## [33,] "0.865384615384615" "0.347826086956522" "linear" "0.125" "2.5"
## [34,] "0.865384615384615" "0.347826086956522" "linear" "0.25" "2.5"
## [35,] "0.865384615384615" "0.347826086956522" "linear" "0.5" "2.5"
## [36,] "0.865384615384615" "0.347826086956522" "linear" "1" "2.5"
## [37,] "0.865384615384615" "0.347826086956522" "linear" "2" "2.5"
## [38,] "0.865384615384615" "0.347826086956522" "linear" "4" "2.5"
## [39,] "0.865384615384615" "0.347826086956522" "linear" "8" "2.5"
## [40,] "0.865384615384615" "0.347826086956522" "linear" "16" "2.5"
## [41,] "0.865384615384615" "0.391304347826087" "linear" "0.125" "3"
## [42,] "0.865384615384615" "0.391304347826087" "linear" "0.25" "3"
## [43,] "0.865384615384615" "0.391304347826087" "linear" "0.5" "3"
## [44,] "0.865384615384615" "0.391304347826087" "linear" "1" "3"
## [45,] "0.865384615384615" "0.391304347826087" "linear" "2" "3"
## [46,] "0.865384615384615" "0.391304347826087" "linear" "4" "3"
## [47,] "0.865384615384615" "0.391304347826087" "linear" "8" "3"
## [48,] "0.865384615384615" "0.391304347826087" "linear" "16" "3"
## [49,] "0.865384615384615" "0.391304347826087" "linear" "0.125" "3.5"
## [50,] "0.865384615384615" "0.391304347826087" "linear" "0.25" "3.5"
## [51,] "0.865384615384615" "0.391304347826087" "linear" "0.5" "3.5"
## [52,] "0.865384615384615" "0.391304347826087" "linear" "1" "3.5"
## [53,] "0.865384615384615" "0.391304347826087" "linear" "2" "3.5"
## [54,] "0.865384615384615" "0.391304347826087" "linear" "4" "3.5"
## [55,] "0.865384615384615" "0.391304347826087" "linear" "8" "3.5"
## [56,] "0.865384615384615" "0.391304347826087" "linear" "16" "3.5"
## [57,] "0.980769230769231" "0.347826086956522" "polynomial" "0.125" ".5"
## [58,] "1" "0.434782608695652" "polynomial" "0.25" ".5"
## [59,] "1" "0.434782608695652" "polynomial" "0.5" ".5"
## [60,] "1" "0.434782608695652" "polynomial" "1" ".5"
## [61,] "1" "0.434782608695652" "polynomial" "2" ".5"
## [62,] "1" "0.434782608695652" "polynomial" "4" ".5"
## [63,] "1" "0.434782608695652" "polynomial" "8" ".5"
## [64,] "1" "0.434782608695652" "polynomial" "16" ".5"
## [65,] "1" "0.434782608695652" "polynomial" "0.125" "1"
## [66,] "1" "0.434782608695652" "polynomial" "0.25" "1"
## [67,] "1" "0.434782608695652" "polynomial" "0.5" "1"
## [68,] "1" "0.434782608695652" "polynomial" "1" "1"
## [69,] "1" "0.434782608695652" "polynomial" "2" "1"
## [70,] "1" "0.434782608695652" "polynomial" "4" "1"
## [71,] "1" "0.434782608695652" "polynomial" "8" "1"
## [72,] "1" "0.434782608695652" "polynomial" "16" "1"
## [73,] "1" "0.434782608695652" "polynomial" "0.125" "1.5"
## [74,] "1" "0.434782608695652" "polynomial" "0.25" "1.5"
## [75,] "1" "0.434782608695652" "polynomial" "0.5" "1.5"
## [76,] "1" "0.434782608695652" "polynomial" "1" "1.5"
## [77,] "1" "0.434782608695652" "polynomial" "2" "1.5"
## [78,] "1" "0.434782608695652" "polynomial" "4" "1.5"
## [79,] "1" "0.434782608695652" "polynomial" "8" "1.5"
## [80,] "1" "0.434782608695652" "polynomial" "16" "1.5"
## [81,] "1" "0.434782608695652" "polynomial" "0.125" "2"
## [82,] "1" "0.434782608695652" "polynomial" "0.25" "2"
## [83,] "1" "0.434782608695652" "polynomial" "0.5" "2"
## [84,] "1" "0.434782608695652" "polynomial" "1" "2"
## [85,] "1" "0.434782608695652" "polynomial" "2" "2"
## [86,] "1" "0.434782608695652" "polynomial" "4" "2"
## [87,] "1" "0.434782608695652" "polynomial" "8" "2"
## [88,] "1" "0.434782608695652" "polynomial" "16" "2"
## [89,] "1" "0.434782608695652" "polynomial" "0.125" "2.5"
## [90,] "1" "0.434782608695652" "polynomial" "0.25" "2.5"
## [91,] "1" "0.434782608695652" "polynomial" "0.5" "2.5"
## [92,] "1" "0.434782608695652" "polynomial" "1" "2.5"
## [93,] "1" "0.434782608695652" "polynomial" "2" "2.5"
## [94,] "1" "0.434782608695652" "polynomial" "4" "2.5"
## [95,] "1" "0.434782608695652" "polynomial" "8" "2.5"
## [96,] "1" "0.434782608695652" "polynomial" "16" "2.5"
## [97,] "1" "0.434782608695652" "polynomial" "0.125" "3"
## [98,] "1" "0.434782608695652" "polynomial" "0.25" "3"
## [99,] "1" "0.434782608695652" "polynomial" "0.5" "3"
## [100,] "1" "0.434782608695652" "polynomial" "1" "3"
## [101,] "1" "0.434782608695652" "polynomial" "2" "3"
## [102,] "1" "0.434782608695652" "polynomial" "4" "3"
## [103,] "1" "0.434782608695652" "polynomial" "8" "3"
## [104,] "1" "0.434782608695652" "polynomial" "16" "3"
## [105,] "1" "0.434782608695652" "polynomial" "0.125" "3.5"
## [106,] "1" "0.434782608695652" "polynomial" "0.25" "3.5"
## [107,] "1" "0.434782608695652" "polynomial" "0.5" "3.5"
## [108,] "1" "0.434782608695652" "polynomial" "1" "3.5"
## [109,] "1" "0.434782608695652" "polynomial" "2" "3.5"
## [110,] "1" "0.434782608695652" "polynomial" "4" "3.5"
## [111,] "1" "0.434782608695652" "polynomial" "8" "3.5"
## [112,] "1" "0.434782608695652" "polynomial" "16" "3.5"
## [113,] "0.615384615384615" "0.608695652173913" "radial" "0.125" ".5"
## [114,] "0.615384615384615" "0.608695652173913" "radial" "0.25" ".5"
## [115,] "0.615384615384615" "0.608695652173913" "radial" "0.5" ".5"
## [116,] "0.615384615384615" "0.608695652173913" "radial" "1" ".5"
## [117,] "0.615384615384615" "0.608695652173913" "radial" "2" ".5"
## [118,] "0.615384615384615" "0.608695652173913" "radial" "4" ".5"
## [119,] "0.615384615384615" "0.608695652173913" "radial" "8" ".5"
## [120,] "0.615384615384615" "0.608695652173913" "radial" "16" ".5"
## [121,] "0.923076923076923" "0.608695652173913" "radial" "0.125" "1"
## [122,] "1" "0.608695652173913" "radial" "0.25" "1"
## [123,] "1" "0.608695652173913" "radial" "0.5" "1"
## [124,] "1" "0.608695652173913" "radial" "1" "1"
## [125,] "1" "0.608695652173913" "radial" "2" "1"
## [126,] "1" "0.608695652173913" "radial" "4" "1"
## [127,] "1" "0.608695652173913" "radial" "8" "1"
## [128,] "1" "0.608695652173913" "radial" "16" "1"
## [129,] "0.961538461538462" "0.565217391304348" "radial" "0.125" "1.5"
## [130,] "1" "0.608695652173913" "radial" "0.25" "1.5"
## [131,] "1" "0.608695652173913" "radial" "0.5" "1.5"
## [132,] "1" "0.608695652173913" "radial" "1" "1.5"
## [133,] "1" "0.608695652173913" "radial" "2" "1.5"
## [134,] "1" "0.608695652173913" "radial" "4" "1.5"
## [135,] "1" "0.608695652173913" "radial" "8" "1.5"
## [136,] "1" "0.608695652173913" "radial" "16" "1.5"
## [137,] "1" "0.521739130434783" "radial" "0.125" "2"
## [138,] "1" "0.652173913043478" "radial" "0.25" "2"
## [139,] "1" "0.608695652173913" "radial" "0.5" "2"
## [140,] "1" "0.608695652173913" "radial" "1" "2"
## [141,] "1" "0.608695652173913" "radial" "2" "2"
## [142,] "1" "0.608695652173913" "radial" "4" "2"
## [143,] "1" "0.608695652173913" "radial" "8" "2"
## [144,] "1" "0.608695652173913" "radial" "16" "2"
## [145,] "1" "0.565217391304348" "radial" "0.125" "2.5"
## [146,] "1" "0.652173913043478" "radial" "0.25" "2.5"
## [147,] "1" "0.608695652173913" "radial" "0.5" "2.5"
## [148,] "1" "0.608695652173913" "radial" "1" "2.5"
## [149,] "1" "0.608695652173913" "radial" "2" "2.5"
## [150,] "1" "0.608695652173913" "radial" "4" "2.5"
## [151,] "1" "0.608695652173913" "radial" "8" "2.5"
## [152,] "1" "0.608695652173913" "radial" "16" "2.5"
## [153,] "1" "0.608695652173913" "radial" "0.125" "3"
## [154,] "1" "0.652173913043478" "radial" "0.25" "3"
## [155,] "1" "0.608695652173913" "radial" "0.5" "3"
## [156,] "1" "0.608695652173913" "radial" "1" "3"
## [157,] "1" "0.608695652173913" "radial" "2" "3"
## [158,] "1" "0.608695652173913" "radial" "4" "3"
## [159,] "1" "0.608695652173913" "radial" "8" "3"
## [160,] "1" "0.608695652173913" "radial" "16" "3"
## [161,] "1" "0.608695652173913" "radial" "0.125" "3.5"
## [162,] "1" "0.652173913043478" "radial" "0.25" "3.5"
## [163,] "1" "0.608695652173913" "radial" "0.5" "3.5"
## [164,] "1" "0.608695652173913" "radial" "1" "3.5"
## [165,] "1" "0.608695652173913" "radial" "2" "3.5"
## [166,] "1" "0.608695652173913" "radial" "4" "3.5"
## [167,] "1" "0.608695652173913" "radial" "8" "3.5"
## [168,] "1" "0.608695652173913" "radial" "16" "3.5"
## [169,] "0.615384615384615" "0.434782608695652" "sigmoid" "0.125" ".5"
## [170,] "0.519230769230769" "0.521739130434783" "sigmoid" "0.25" ".5"
## [171,] "0.519230769230769" "0.521739130434783" "sigmoid" "0.5" ".5"
## [172,] "0.461538461538462" "0.434782608695652" "sigmoid" "1" ".5"
## [173,] "0.442307692307692" "0.434782608695652" "sigmoid" "2" ".5"
## [174,] "0.461538461538462" "0.521739130434783" "sigmoid" "4" ".5"
## [175,] "0.442307692307692" "0.608695652173913" "sigmoid" "8" ".5"
## [176,] "0.442307692307692" "0.608695652173913" "sigmoid" "16" ".5"
## [177,] "0.615384615384615" "0.521739130434783" "sigmoid" "0.125" "1"
## [178,] "0.5" "0.521739130434783" "sigmoid" "0.25" "1"
## [179,] "0.423076923076923" "0.565217391304348" "sigmoid" "0.5" "1"
## [180,] "0.423076923076923" "0.565217391304348" "sigmoid" "1" "1"
## [181,] "0.442307692307692" "0.652173913043478" "sigmoid" "2" "1"
## [182,] "0.442307692307692" "0.521739130434783" "sigmoid" "4" "1"
## [183,] "0.461538461538462" "0.521739130434783" "sigmoid" "8" "1"
## [184,] "0.442307692307692" "0.521739130434783" "sigmoid" "16" "1"
## [185,] "0.538461538461538" "0.478260869565217" "sigmoid" "0.125" "1.5"
## [186,] "0.5" "0.521739130434783" "sigmoid" "0.25" "1.5"
## [187,] "0.442307692307692" "0.565217391304348" "sigmoid" "0.5" "1.5"
## [188,] "0.423076923076923" "0.608695652173913" "sigmoid" "1" "1.5"
## [189,] "0.461538461538462" "0.565217391304348" "sigmoid" "2" "1.5"
## [190,] "0.5" "0.565217391304348" "sigmoid" "4" "1.5"
## [191,] "0.5" "0.652173913043478" "sigmoid" "8" "1.5"
## [192,] "0.442307692307692" "0.608695652173913" "sigmoid" "16" "1.5"
## [193,] "0.5" "0.478260869565217" "sigmoid" "0.125" "2"
## [194,] "0.461538461538462" "0.521739130434783" "sigmoid" "0.25" "2"
## [195,] "0.442307692307692" "0.565217391304348" "sigmoid" "0.5" "2"
## [196,] "0.5" "0.478260869565217" "sigmoid" "1" "2"
## [197,] "0.461538461538462" "0.565217391304348" "sigmoid" "2" "2"
## [198,] "0.519230769230769" "0.565217391304348" "sigmoid" "4" "2"
## [199,] "0.519230769230769" "0.652173913043478" "sigmoid" "8" "2"
## [200,] "0.442307692307692" "0.608695652173913" "sigmoid" "16" "2"
## [201,] "0.519230769230769" "0.434782608695652" "sigmoid" "0.125" "2.5"
## [202,] "0.461538461538462" "0.521739130434783" "sigmoid" "0.25" "2.5"
## [203,] "0.442307692307692" "0.565217391304348" "sigmoid" "0.5" "2.5"
## [204,] "0.461538461538462" "0.434782608695652" "sigmoid" "1" "2.5"
## [205,] "0.519230769230769" "0.608695652173913" "sigmoid" "2" "2.5"
## [206,] "0.519230769230769" "0.565217391304348" "sigmoid" "4" "2.5"
## [207,] "0.519230769230769" "0.652173913043478" "sigmoid" "8" "2.5"
## [208,] "0.442307692307692" "0.608695652173913" "sigmoid" "16" "2.5"
## [209,] "0.5" "0.434782608695652" "sigmoid" "0.125" "3"
## [210,] "0.461538461538462" "0.521739130434783" "sigmoid" "0.25" "3"
## [211,] "0.442307692307692" "0.565217391304348" "sigmoid" "0.5" "3"
## [212,] "0.442307692307692" "0.434782608695652" "sigmoid" "1" "3"
## [213,] "0.538461538461538" "0.608695652173913" "sigmoid" "2" "3"
## [214,] "0.519230769230769" "0.565217391304348" "sigmoid" "4" "3"
## [215,] "0.557692307692308" "0.521739130434783" "sigmoid" "8" "3"
## [216,] "0.442307692307692" "0.608695652173913" "sigmoid" "16" "3"
## [217,] "0.5" "0.434782608695652" "sigmoid" "0.125" "3.5"
## [218,] "0.461538461538462" "0.565217391304348" "sigmoid" "0.25" "3.5"
## [219,] "0.442307692307692" "0.565217391304348" "sigmoid" "0.5" "3.5"
## [220,] "0.442307692307692" "0.434782608695652" "sigmoid" "1" "3.5"
## [221,] "0.442307692307692" "0.565217391304348" "sigmoid" "2" "3.5"
## [222,] "0.519230769230769" "0.565217391304348" "sigmoid" "4" "3.5"
## [223,] "0.557692307692308" "0.521739130434783" "sigmoid" "8" "3.5"
## [224,] "0.442307692307692" "0.608695652173913" "sigmoid" "16" "3.5"# run model using best cost, gamma, kernel and cross-validation
set.seed(321)
svm_model_cs_tuned <- svm(SPSyn~ ., data=training_set, method="C-classification", kernel="sigmoid",gamma=.125,cost=1, cross=3)
#training set predictions
pred_train <-predict(svm_model_cs_tuned,training_set)
mean(pred_train==training_set$SPSyn)## [1] 0.6153846confusionMatrix(pred_train,training_set$SPSyn)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 24 12
## 1 8 8
##
## Accuracy : 0.6154
## 95% CI : (0.4702, 0.747)
## No Information Rate : 0.6154
## P-Value [Acc > NIR] : 0.5608
##
## Kappa : 0.1558
##
## Mcnemar's Test P-Value : 0.5023
##
## Sensitivity : 0.7500
## Specificity : 0.4000
## Pos Pred Value : 0.6667
## Neg Pred Value : 0.5000
## Prevalence : 0.6154
## Detection Rate : 0.4615
## Detection Prevalence : 0.6923
## Balanced Accuracy : 0.5750
##
## 'Positive' Class : 0
## #test set predictions
pred_test <-predict(svm_model_cs_tuned,test_set)
mean(pred_test==test_set$SPSyn)## [1] 0.5217391confusionMatrix(pred_test,test_set$SPSyn)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 9 6
## 1 5 3
##
## Accuracy : 0.5217
## 95% CI : (0.3059, 0.7318)
## No Information Rate : 0.6087
## P-Value [Acc > NIR] : 0.857
##
## Kappa : -0.0243
##
## Mcnemar's Test P-Value : 1.000
##
## Sensitivity : 0.6429
## Specificity : 0.3333
## Pos Pred Value : 0.6000
## Neg Pred Value : 0.3750
## Prevalence : 0.6087
## Detection Rate : 0.3913
## Detection Prevalence : 0.6522
## Balanced Accuracy : 0.4881
##
## 'Positive' Class : 0
## K Nearest Neighbors
set.seed(321)
split = sample.split(temat.cog.fs.flux.2.imputed$SPSyn, SplitRatio = 0.70)
training_set = subset(temat.cog.fs.flux.2.imputed, split == TRUE)
test_set = subset(temat.cog.fs.flux.2.imputed, split == FALSE)
training_set$SPSyn <- as.factor(training_set$SPSyn)
test_set$SPSyn <- as.factor(test_set$SPSyn)
#feature scaling
training_set[-22] = scale(training_set[-22])
test_set[-22] = scale(test_set[-22])
#Model tutning
set.seed(7)
ctrl <- trainControl(method="cv",number = 3) #,classProbs=TRUE,summaryFunction = twoClassSummary)
knnFit <- train(SPSyn ~ ., data = training_set, method = "knn", trControl = ctrl, tuneLength = 3) # optimal k = 5
knnFit## k-Nearest Neighbors
##
## 52 samples
## 21 predictors
## 2 classes: '0', '1'
##
## No pre-processing
## Resampling: Cross-Validated (3 fold)
## Summary of sample sizes: 34, 35, 35
## Resampling results across tuning parameters:
##
## k Accuracy Kappa
## 5 0.5969499 -0.01449275
## 7 0.5784314 -0.02070494
## 9 0.5773420 -0.05187593
##
## Accuracy was used to select the optimal model using the largest value.
## The final value used for the model was k = 5.#training set predictions
pred_train_knn <-predict(knnFit,training_set)
mean(pred_train_knn==training_set$SPSyn)## [1] 0.6730769confusionMatrix(pred_train_knn,training_set$SPSyn)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 28 13
## 1 4 7
##
## Accuracy : 0.6731
## 95% CI : (0.5289, 0.7967)
## No Information Rate : 0.6154
## P-Value [Acc > NIR] : 0.23993
##
## Kappa : 0.2457
##
## Mcnemar's Test P-Value : 0.05235
##
## Sensitivity : 0.8750
## Specificity : 0.3500
## Pos Pred Value : 0.6829
## Neg Pred Value : 0.6364
## Prevalence : 0.6154
## Detection Rate : 0.5385
## Detection Prevalence : 0.7885
## Balanced Accuracy : 0.6125
##
## 'Positive' Class : 0
## #test set predictions
pred_test_knn <-predict(knnFit,test_set)
mean(pred_test_knn==test_set$SPSyn)## [1] 0.5217391confusionMatrix(pred_test_knn,test_set$SPSyn)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 12 9
## 1 2 0
##
## Accuracy : 0.5217
## 95% CI : (0.3059, 0.7318)
## No Information Rate : 0.6087
## P-Value [Acc > NIR] : 0.85697
##
## Kappa : -0.1659
##
## Mcnemar's Test P-Value : 0.07044
##
## Sensitivity : 0.8571
## Specificity : 0.0000
## Pos Pred Value : 0.5714
## Neg Pred Value : 0.0000
## Prevalence : 0.6087
## Detection Rate : 0.5217
## Detection Prevalence : 0.9130
## Balanced Accuracy : 0.4286
##
## 'Positive' Class : 0
##