library(e1071)
library(readr)
library(caret)
## Loading required package: lattice
## Loading required package: ggplot2
library(MASS)
Records <- read.csv("C:/Users/aksha/Desktop/Shiny app/RPV/Arr - noshow RPV.csv",sep=',')
#View(Records)
#Removing NA's from zip column
Records = Records[complete.cases(Records[ ,2]),]
str(Records)
## 'data.frame': 9049 obs. of 56 variables:
## $ ï..MRN : int 5169984 5169984 5169984 5169984 5169984 5169984 5169984 2908595 2929531 5169984 ...
## $ ZIP : Factor w/ 446 levels "07001","07002",..: 133 133 133 133 133 133 133 391 359 133 ...
## $ DistanceToClinic : num 34.3 34.3 34.3 34.3 34.3 34.3 34.3 0.8 10.6 34.3 ...
## $ AGE : int 60 60 60 60 60 60 60 63 61 60 ...
## $ LOCATION : Factor w/ 1 level "CAB 6TH FLOOR": 1 1 1 1 1 1 1 1 1 1 ...
## $ DT : Factor w/ 494 levels "1/10/2014","1/12/2015",..: 107 355 122 219 371 383 452 36 109 196 ...
## $ Time : Factor w/ 45 levels "01:00PM","01:10PM",..: 29 27 29 29 31 29 27 44 33 33 ...
## $ TimeFrame_Hour : int 9 8 9 9 9 9 8 12 10 10 ...
## $ Weekday : Factor w/ 5 levels "Friday","Monday",..: 2 3 2 3 2 5 3 3 4 5 ...
## $ Month : Factor w/ 12 levels "April","August",..: 10 7 3 8 6 6 2 5 10 4 ...
## $ Season : logi NA NA NA NA NA NA ...
## $ SCHED.PROV : Factor w/ 25 levels "AHMAD,HAROON RES",..: 3 3 3 21 3 3 3 12 3 3 ...
## $ SCHEDPROV_LastName : Factor w/ 25 levels "AHMAD","ALBRECHT",..: 3 3 3 20 3 3 3 11 3 3 ...
## $ VT : Factor w/ 11 levels "BTR","DOP","MDR",..: 5 5 5 5 8 8 8 11 4 5 ...
## $ VisitType : Factor w/ 1 level "RPV": 1 1 1 1 1 1 1 1 1 1 ...
## $ DURATION : int 60 60 60 60 60 60 60 45 30 60 ...
## $ CANCEL.DT : logi NA NA NA NA NA NA ...
## $ CAN.BUMP.INITIAL : logi NA NA NA NA NA NA ...
## $ CAN.BUMP.INITITALS : logi NA NA NA NA NA NA ...
## $ CANCEL.REASON : logi NA NA NA NA NA NA ...
## $ PCC : Factor w/ 1515 levels "","ABBAS,SHAHIDA M",..: 1434 1434 1434 1434 1434 1434 1434 1200 1393 1434 ...
## $ Lead.Time : int 0 0 0 0 0 0 0 0 1 1 ...
## $ DT.WHEN.SCHED : Factor w/ 600 levels "1/10/2014","1/12/2015",..: 118 452 133 253 473 485 557 47 118 216 ...
## $ DT.WHEN.RESCHED : logi NA NA NA NA NA NA ...
## $ COMMENTS : Factor w/ 1028 levels "1 MON F/U","1 YEAR F/U (DR WONG)",..: 246 246 246 246 302 320 312 121 204 241 ...
## $ MARITAL : Factor w/ 5 levels "DIVORCED","MARRIED",..: 2 2 2 2 2 2 2 2 2 2 ...
## $ SEX : Factor w/ 2 levels "F","M": 2 2 2 2 2 2 2 2 2 2 ...
## $ EMPLOYER : Factor w/ 15 levels "","AT&T","HAIMM,NEIL",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ EMPLOYER.GROUP : logi NA NA NA NA NA NA ...
## $ REG.FSC : Factor w/ 52 levels "AETNA HMO","AETNA MEDICARE",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ REG.FSC.1 : Factor w/ 5 levels "Commercial","Indigent",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ SCH.PROV.CATEGORY : Factor w/ 4 levels "EPILEPSY","GEN NEUROLOGY/HEADACHE",..: 3 3 3 3 3 3 3 4 3 3 ...
## $ SCH.PROV.CATEGORY.1 : Factor w/ 4 levels "EPILEPSY","GENNEUROLOGY_HEADACHE",..: 3 3 3 3 3 3 3 4 3 3 ...
## $ INV..BILLED : int 11426195 12223483 12434701 11807074 11426178 11426182 11426187 12403675 11525145 12868203 ...
## $ INVBAL : num 0 0 0 0 0 0 0 0 0 5 ...
## $ invoicebalance : num 0 0 0 0 0 0 0 0 0 5 ...
## $ BILLING.PROVIDER : Factor w/ 20 levels "","ALBRECHT,CATHERINE",..: 18 18 4 18 4 4 18 18 18 4 ...
## $ SERVICING.PROVIDER : Factor w/ 33 levels "","AHMAD,HAROON RES",..: 5 5 5 29 5 5 5 17 5 5 ...
## $ HOS : Factor w/ 5 levels "","CANCER INSTITUTE OF NEW JERSEY",..: 4 4 4 4 4 4 4 4 4 4 ...
## $ DX1 : Factor w/ 241 levels "","000","000.0",..: 64 64 64 64 64 64 64 124 64 64 ...
## $ DX1.DESCRIPTION : Factor w/ 241 levels "","ABNORMAL FINDINGS SEMEN",..: 161 161 161 161 161 161 161 223 161 161 ...
## $ DX2 : Factor w/ 325 levels "","053.19","078.5",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DX2.DESCRIPTION : Factor w/ 325 levels "","ABDOM/PELVIC SWELLING UNSP SITE",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DX3 : Factor w/ 244 levels "","013.04","042",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DX3.DESCRIPTION : Factor w/ 244 levels "","ABNORMAL CNS FUNCT STUDY OT",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DX4 : Factor w/ 136 levels "","183.0","209.29",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DX4.DESCRIPTION : Factor w/ 136 levels "","ABNORM EXAM FINDINGS,OTHER",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ DX5 : num NA NA NA NA NA NA NA NA NA NA ...
## $ DX5.DESCRIPTION : Factor w/ 12 levels "","DISPLACE INTERVERT DISC SITE UNS",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ REF.PROV : Factor w/ 1610 levels "","ABBAS,SHAHIDA M",..: 1524 1524 1524 1524 1524 1524 1524 1271 1482 1524 ...
## $ REF.PROV.ZIP : Factor w/ 300 levels "","01812","06512",..: 204 204 204 204 204 204 204 255 218 204 ...
## $ REF.PROV.SPEC : logi NA NA NA NA NA NA ...
## $ STATUS : Factor w/ 2 levels "ARR","NOS": 1 1 1 1 1 1 1 1 1 1 ...
## $ CancellationTiming_Days: Factor w/ 2 levels "No Can","No Show": 1 1 1 1 1 1 1 1 1 1 ...
## $ CancellationCategory : Factor w/ 1 level "0hr": 1 1 1 1 1 1 1 1 1 1 ...
## $ Outcome : Factor w/ 2 levels "ARR","NOS": 1 1 1 1 1 1 1 1 1 1 ...
summary(Records)
## ï..MRN ZIP DistanceToClinic AGE
## Min. :2902470 08831 : 500 Min. : 0.80 Min. :17.00
## 1st Qu.:2992111 08873 : 359 1st Qu.: 8.10 1st Qu.:50.00
## Median :3623995 08901 : 326 Median : 14.50 Median :64.00
## Mean :4013417 08816 : 268 Mean : 20.96 Mean :60.76
## 3rd Qu.:5145983 08854 : 245 3rd Qu.: 32.60 3rd Qu.:73.00
## Max. :5436187 08902 : 238 Max. :120.00 Max. :96.00
## (Other):7113
## LOCATION DT Time TimeFrame_Hour
## CAB 6TH FLOOR:9049 3/16/2015 : 42 12:30PM: 977 Min. : 1.000
## 12/15/2014: 39 11:00AM: 830 1st Qu.: 3.000
## 3/30/2015 : 38 10:00AM: 782 Median :10.000
## 8/18/2014 : 38 10:30AM: 680 Mean : 8.013
## 9/9/2013 : 37 09:00AM: 653 3rd Qu.:11.000
## 1/12/2015 : 34 09:30AM: 562 Max. :12.000
## (Other) :8821 (Other):4565
## Weekday Month Season
## Friday : 964 October : 868 Mode:logical
## Monday :2208 May : 856 NA's:9049
## Thursday :2287 March : 838
## Tuesday :2096 April : 799
## Wednesday:1494 September: 791
## December : 737
## (Other) :4160
## SCHED.PROV SCHEDPROV_LastName VT VisitType
## MARK,MARGERY :1527 MARK :1527 RPV :6286 RPV:9049
## SAGE,JACOB :1370 SAGE :1370 BTR :1292
## GOLBE,LAWRENCE :1214 GOLBE :1214 P60 : 550
## SCHNEIDER,DANIEL :1113 SCHNEIDER:1113 RBH : 406
## CAPUTO,DEBORAH : 836 CAPUTO : 836 PSP : 198
## ALBRECHT,CATHERINE: 638 ALBRECHT : 638 P30 : 142
## (Other) :2351 (Other) :2351 (Other): 175
## DURATION CANCEL.DT CAN.BUMP.INITIAL CAN.BUMP.INITITALS
## Min. : 15.00 Mode:logical Mode:logical Mode:logical
## 1st Qu.: 30.00 NA's:9049 NA's:9049 NA's:9049
## Median : 30.00
## Mean : 32.34
## 3rd Qu.: 30.00
## Max. :180.00
##
## CANCEL.REASON PCC Lead.Time DT.WHEN.SCHED
## Mode:logical : 159 Min. : 0.0 12/1/2014 : 50
## NA's:9049 HASTINGS,SHIRIN: 91 1st Qu.: 14.0 9/17/2013 : 45
## YU,FRAN : 86 Median : 45.0 2/2/2015 : 44
## OTHER,REFPHYS : 74 Mean : 63.2 1/26/2015 : 38
## ROSENFELD,JANE : 64 3rd Qu.: 98.0 12/15/2014: 38
## ARMAS,BARBARA J: 60 Max. :205.0 10/29/2013: 37
## (Other) :8515 (Other) :8797
## DT.WHEN.RESCHED COMMENTS MARITAL SEX
## Mode:logical RPV :3655 DIVORCED : 555 F:4435
## NA's:9049 BTR :1081 MARRIED :5339 M:4614
## RPV/FOLLOW UP: 833 SEPARATED: 100
## RBH : 290 SINGLE :2359
## F/U : 258 WIDOWED : 696
## P60 : 203
## (Other) :2729
## EMPLOYER EMPLOYER.GROUP
## :8994 Mode:logical
## MIDD CTY BD OF SOC SVCS: 7 NA's:9049
## RETIRED : 6
## HAIMM,NEIL : 5
## phil : 5
## RWJ : 5
## (Other) : 27
## REG.FSC REG.FSC.1
## MEDICARE US :4578 Commercial:2616
## HORIZON PPO : 806 Indigent : 224
## HORIZON NJ HEALTH HORIZON MCAID: 588 Medicaid :1145
## HORIZON POS : 484 Medicare :5057
## UNITED HEALTHCARE MEDICAID : 461 Other : 7
## AETNA PPO : 250
## (Other) :1882
## SCH.PROV.CATEGORY SCH.PROV.CATEGORY.1
## EPILEPSY :1465 EPILEPSY :1465
## GEN NEUROLOGY/HEADACHE: 912 GENNEUROLOGY_HEADACHE: 912
## MOVEMENT DISORDERS :6060 MOVEMENT DISORDERS :6060
## RESIDENT : 612 RESIDENT : 612
##
##
##
## INV..BILLED INVBAL invoicebalance
## Min. :11004964 Min. :-310.22 Min. :-310.22
## 1st Qu.:11523379 1st Qu.: 0.00 1st Qu.: 0.00
## Median :11980990 Median : 0.00 Median : 0.00
## Mean :11968257 Mean : 13.38 Mean : 13.38
## 3rd Qu.:12428877 3rd Qu.: 0.00 3rd Qu.: 0.00
## Max. :13092460 Max. :8456.00 Max. :8456.00
## NA's :843 NA's :843 NA's :843
## BILLING.PROVIDER SERVICING.PROVIDER
## MARK,MARGERY :1589 MARK,MARGERY :1452
## SCHNEIDER,DANIEL:1499 SAGE,JACOB :1300
## SAGE,JACOB :1300 GOLBE,LAWRENCE :1141
## GOLBE,LAWRENCE :1142 SCHNEIDER,DANIEL:1015
## : 843 : 843
## MANI,RAM : 487 CAPUTO,DEBORAH : 792
## (Other) :2189 (Other) :2506
## HOS DX1
## : 845 332.0 :3440
## CANCER INSTITUTE OF NEW JERSEY : 14 : 843
## CHILD HEALTH INSTITUE OF NEW JE: 5 333.83 : 607
## CLINICAL ACADEMIC BUILDING :8174 345.41 : 373
## ROBERT WOOD JOHNSON HOSPITAL : 11 784.0 : 313
## 351.8 : 302
## (Other):3171
## DX1.DESCRIPTION DX2
## PARALYSIS AGITANS :3440 :7559
## : 843 784.0 : 109
## SPASMODIC TORTICOLLIS : 607 332.0 : 97
## PARTIAL EPILEPSY IMPAIRMENT INTRAC: 373 333.83 : 60
## HEADACHE : 313 780.93 : 48
## OTH FACIAL NERVE DISORDERS : 302 723.1 : 43
## (Other) :3171 (Other):1133
## DX2.DESCRIPTION DX3
## :7559 :8398
## HEADACHE : 109 356.9 : 32
## PARALYSIS AGITANS : 97 784.0 : 30
## SPASMODIC TORTICOLLIS: 60 780.93 : 21
## MEMORY LOSS : 48 782.0 : 18
## CERVICALGIA : 43 332.0 : 15
## (Other) :1133 (Other): 535
## DX3.DESCRIPTION DX4
## :8398 :8726
## UNS IDIOPATHIC PERIPH NEUROPATHY: 32 V26.33 : 15
## HEADACHE : 30 356.9 : 13
## MEMORY LOSS : 21 724.5 : 9
## DISTURBANCE SKIN SENSATION : 18 781.2 : 8
## PARALYSIS AGITANS : 15 784.0 : 8
## (Other) : 535 (Other): 270
## DX4.DESCRIPTION DX5
## :8726 Min. : 93.89
## GENETIC COUNSELING : 15 1st Qu.:266.00
## UNS IDIOPATHIC PERIPH NEUROPATHY: 13 Median :357.86
## BACKACHE UNSPECIFIED : 9 Mean :445.65
## ABNORMALITY OF GAIT : 8 3rd Qu.:736.88
## HEADACHE : 8 Max. :787.20
## (Other) : 270 NA's :9037
## DX5.DESCRIPTION REF.PROV
## :9037 :1266
## MEMORY LOSS : 2 MARK,MARGERY H : 78
## DISPLACE INTERVERT DISC SITE UNS : 1 ARMAS,BARBARA J: 70
## DYSPHAGIA,UNSPECIFIED : 1 ROSENFELD,JANE : 62
## OTH BENIGN NEO CONNEC SOFT TISS UN: 1 KIM,SARANG : 57
## OTHER CARDIOVASCULAR SYPHILIS : 1 YU,FRAN : 48
## (Other) : 6 (Other) :7468
## REF.PROV.ZIP REF.PROV.SPEC STATUS CancellationTiming_Days
## :1898 Mode:logical ARR:8227 No Can :8227
## 08901 :1062 NA's:9049 NOS: 822 No Show: 822
## 08816 : 342
## 08903 : 223
## 08831 : 222
## 08820 : 198
## (Other):5104
## CancellationCategory Outcome
## 0hr:9049 ARR:8227
## NOS: 822
##
##
##
##
##
#Conveting to nominal and numeric attributes
Records$ZIP = as.numeric(Records$ZIP)
Records$DistanceToClinic = as.integer(Records$DistanceToClinic)
Records$TimeFrame_Hour = as.factor(Records$TimeFrame_Hour)
Records$Weekday = as.factor(Records$Weekday)
Records$Month = as.factor(Records$Month)
Records$SCHED.PROV = as.factor(Records$SCHED.PROV)
Records$VT = as.factor(Records$VT)
Records$VisitType = as.factor(Records$VisitType)
Records$MARITAL = as.factor(Records$MARITAL)
Records$SEX = as.factor(Records$SEX)
Records$REG.FSC.1 = as.factor(Records$REG.FSC.1)
Records$SCH.PROV.CATEGORY.1 = as.factor(Records$SCH.PROV.CATEGORY.1)
Records$Outcome = as.factor(Records$Outcome)
Records$STATUS = as.factor(Records$STATUS)
myvars <- c("Outcome", "DistanceToClinic","SEX","MARITAL","AGE","TimeFrame_Hour","Weekday","Month","DURATION","REG.FSC.1","SCH.PROV.CATEGORY.1")
newdata <- Records[myvars]
newdata$AGE = as.numeric(newdata$AGE)
newdata$DURATION = as.numeric(newdata$DURATION)
summary(newdata)
## Outcome DistanceToClinic SEX MARITAL AGE
## ARR:8227 Min. : 0.00 F:4435 DIVORCED : 555 Min. :17.00
## NOS: 822 1st Qu.: 8.00 M:4614 MARRIED :5339 1st Qu.:50.00
## Median : 14.00 SEPARATED: 100 Median :64.00
## Mean : 20.48 SINGLE :2359 Mean :60.76
## 3rd Qu.: 32.00 WIDOWED : 696 3rd Qu.:73.00
## Max. :120.00 Max. :96.00
##
## TimeFrame_Hour Weekday Month DURATION
## 12 :1559 Friday : 964 October : 868 Min. : 15.00
## 10 :1543 Monday :2208 May : 856 1st Qu.: 30.00
## 11 :1476 Thursday :2287 March : 838 Median : 30.00
## 9 :1320 Tuesday :2096 April : 799 Mean : 32.34
## 1 :1115 Wednesday:1494 September: 791 3rd Qu.: 30.00
## 2 : 721 December : 737 Max. :180.00
## (Other):1315 (Other) :4160
## REG.FSC.1 SCH.PROV.CATEGORY.1
## Commercial:2616 EPILEPSY :1465
## Indigent : 224 GENNEUROLOGY_HEADACHE: 912
## Medicaid :1145 MOVEMENT DISORDERS :6060
## Medicare :5057 RESIDENT : 612
## Other : 7
##
##
str(newdata)
## 'data.frame': 9049 obs. of 11 variables:
## $ Outcome : Factor w/ 2 levels "ARR","NOS": 1 1 1 1 1 1 1 1 1 1 ...
## $ DistanceToClinic : int 34 34 34 34 34 34 34 0 10 34 ...
## $ SEX : Factor w/ 2 levels "F","M": 2 2 2 2 2 2 2 2 2 2 ...
## $ MARITAL : Factor w/ 5 levels "DIVORCED","MARRIED",..: 2 2 2 2 2 2 2 2 2 2 ...
## $ AGE : num 60 60 60 60 60 60 60 63 61 60 ...
## $ TimeFrame_Hour : Factor w/ 10 levels "1","2","3","4",..: 7 6 7 7 7 7 6 10 8 8 ...
## $ Weekday : Factor w/ 5 levels "Friday","Monday",..: 2 3 2 3 2 5 3 3 4 5 ...
## $ Month : Factor w/ 12 levels "April","August",..: 10 7 3 8 6 6 2 5 10 4 ...
## $ DURATION : num 60 60 60 60 60 60 60 45 30 60 ...
## $ REG.FSC.1 : Factor w/ 5 levels "Commercial","Indigent",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ SCH.PROV.CATEGORY.1: Factor w/ 4 levels "EPILEPSY","GENNEUROLOGY_HEADACHE",..: 3 3 3 3 3 3 3 4 3 3 ...
set.seed(5000)
samples <- sample(nrow(newdata),as.integer(nrow(newdata)*0.75))
train.newdata = newdata[samples,]
test.newdata = newdata[-samples,]
#1) SVM Classification
library(e1071)
model1<-svm(Outcome ~ DistanceToClinic + SEX + MARITAL + AGE + TimeFrame_Hour +Weekday + Month + DURATION + REG.FSC.1 + SCH.PROV.CATEGORY.1, data = train.newdata)
#Summarize the model
summary(model1)
##
## Call:
## svm(formula = Outcome ~ DistanceToClinic + SEX + MARITAL + AGE +
## TimeFrame_Hour + Weekday + Month + DURATION + REG.FSC.1 +
## SCH.PROV.CATEGORY.1, data = train.newdata)
##
##
## Parameters:
## SVM-Type: C-classification
## SVM-Kernel: radial
## cost: 1
## gamma: 0.025
##
## Number of Support Vectors: 2176
##
## ( 1575 601 )
##
##
## Number of Classes: 2
##
## Levels:
## ARR NOS
#Predict using the model
pred_model1 = predict(model1,test.newdata,type="response")
mtab_model1<-table(pred_model1,test.newdata$Outcome)
confusionMatrix(mtab_model1)
## Confusion Matrix and Statistics
##
##
## pred_model1 ARR NOS
## ARR 2042 221
## NOS 0 0
##
## Accuracy : 0.9023
## 95% CI : (0.8894, 0.9143)
## No Information Rate : 0.9023
## P-Value [Acc > NIR] : 0.5179
##
## Kappa : 0
## Mcnemar's Test P-Value : <2e-16
##
## Sensitivity : 1.0000
## Specificity : 0.0000
## Pos Pred Value : 0.9023
## Neg Pred Value : NaN
## Prevalence : 0.9023
## Detection Rate : 0.9023
## Detection Prevalence : 1.0000
## Balanced Accuracy : 0.5000
##
## 'Positive' Class : ARR
##
accuracy_75_model1_no_show_Arrival_RPV = sum(diag(table(test.newdata$Outcome,pred_model1)))/nrow(test.newdata)
accuracy_75_model1_no_show_Arrival_RPV
## [1] 0.902342
#2) LOGISTIC Regression
model2<-glm(Outcome ~ DistanceToClinic + SEX + MARITAL + AGE + TimeFrame_Hour +Weekday + Month + DURATION + REG.FSC.1 + SCH.PROV.CATEGORY.1, data = train.newdata, family = "binomial" )
pred1 = predict(model2,test.newdata,type="response")
pred1=ifelse(pred1<0.5,0,1)
accuracy_75_log_arrival_noshow_rpv = sum(diag(table(test.newdata$Outcome,pred1)))/nrow(test.newdata)
accuracy_75_log_arrival_noshow_rpv
## [1] 0.902342
#3) STEP REGRESSION
model1_null<-glm(Outcome ~ 1, data = train.newdata,family = "binomial")
model1_all<-glm(Outcome ~ ., data = train.newdata,family = "binomial")
forward_model = stepAIC(model1_null, direction='forward', scope=list(lower=model1_null,upper=model1_all))
## Start: AIC=4062.8
## Outcome ~ 1
##
## Df Deviance AIC
## + SCH.PROV.CATEGORY.1 3 3859.2 3867.2
## + AGE 1 3915.6 3919.6
## + REG.FSC.1 4 3951.5 3961.5
## + MARITAL 4 3957.2 3967.2
## + DistanceToClinic 1 4029.8 4033.8
## + TimeFrame_Hour 9 4035.5 4055.5
## + Weekday 4 4047.2 4057.2
## + SEX 1 4053.4 4057.4
## <none> 4060.8 4062.8
## + DURATION 1 4060.8 4064.8
## + Month 11 4046.1 4070.1
##
## Step: AIC=3867.19
## Outcome ~ SCH.PROV.CATEGORY.1
##
## Df Deviance AIC
## + AGE 1 3835.9 3845.9
## + MARITAL 4 3839.4 3855.4
## + REG.FSC.1 4 3841.7 3857.7
## + DURATION 1 3851.2 3861.2
## + DistanceToClinic 1 3856.7 3866.7
## <none> 3859.2 3867.2
## + Weekday 4 3851.6 3867.6
## + SEX 1 3858.0 3868.0
## + Month 11 3846.6 3876.6
## + TimeFrame_Hour 9 3854.0 3880.0
##
## Step: AIC=3845.88
## Outcome ~ SCH.PROV.CATEGORY.1 + AGE
##
## Df Deviance AIC
## + DURATION 1 3826.5 3838.5
## + MARITAL 4 3824.1 3842.1
## + REG.FSC.1 4 3825.4 3843.4
## + DistanceToClinic 1 3832.5 3844.5
## <none> 3835.9 3845.9
## + SEX 1 3834.5 3846.5
## + Weekday 4 3829.2 3847.2
## + Month 11 3823.1 3855.1
## + TimeFrame_Hour 9 3829.8 3857.8
##
## Step: AIC=3838.46
## Outcome ~ SCH.PROV.CATEGORY.1 + AGE + DURATION
##
## Df Deviance AIC
## + MARITAL 4 3814.4 3834.4
## + REG.FSC.1 4 3816.5 3836.5
## + DistanceToClinic 1 3823.6 3837.6
## <none> 3826.5 3838.5
## + SEX 1 3825.4 3839.4
## + Weekday 4 3821.6 3841.6
## + Month 11 3814.7 3848.7
## + TimeFrame_Hour 9 3819.0 3849.0
##
## Step: AIC=3834.45
## Outcome ~ SCH.PROV.CATEGORY.1 + AGE + DURATION + MARITAL
##
## Df Deviance AIC
## + DistanceToClinic 1 3812.0 3834.0
## <none> 3814.4 3834.4
## + REG.FSC.1 4 3807.6 3835.6
## + SEX 1 3814.1 3836.1
## + Weekday 4 3809.1 3837.1
## + Month 11 3802.4 3844.4
## + TimeFrame_Hour 9 3807.0 3845.0
##
## Step: AIC=3834
## Outcome ~ SCH.PROV.CATEGORY.1 + AGE + DURATION + MARITAL + DistanceToClinic
##
## Df Deviance AIC
## <none> 3812.0 3834.0
## + REG.FSC.1 4 3804.6 3834.6
## + SEX 1 3811.6 3835.6
## + Weekday 4 3807.1 3837.1
## + Month 11 3800.1 3844.1
## + TimeFrame_Hour 9 3804.4 3844.4
pred1 = predict(forward_model,test.newdata,type="response")
pred1=ifelse(pred1<0.5,0,1)
accuracy_75_step_arrival_noshow_rpv = sum(diag(table(test.newdata$Outcome,pred1)))/nrow(test.newdata)
accuracy_75_step_arrival_noshow_rpv
## [1] 0.902342