Goal is to automate building and tuning a classification model to predict employee attrition, using the h2o::h2o.automl.
Import the cleaned data from Module 7.
library(h2o)##
## ----------------------------------------------------------------------
##
## Your next step is to start H2O:
## > h2o.init()
##
## For H2O package documentation, ask for help:
## > ??h2o
##
## After starting H2O, you can use the Web UI at http://localhost:54321
## For more information visit https://docs.h2o.ai
##
## ----------------------------------------------------------------------##
## Attaching package: 'h2o'## The following objects are masked from 'package:stats':
##
## cor, sd, var## The following objects are masked from 'package:base':
##
## &&, %*%, %in%, ||, apply, as.factor, as.numeric, colnames,
## colnames<-, ifelse, is.character, is.factor, is.numeric, log,
## log10, log1p, log2, round, signif, trunclibrary(tidyverse)## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2
## ──## ✔ ggplot2 3.4.4 ✔ purrr 1.0.2
## ✔ tibble 3.2.1 ✔ dplyr 1.1.4
## ✔ tidyr 1.3.0 ✔ stringr 1.5.0
## ✔ readr 2.1.3 ✔ forcats 1.0.0
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()library(tidymodels)## ── Attaching packages ────────────────────────────────────── tidymodels 1.1.1 ──
## ✔ broom 1.0.5 ✔ rsample 1.2.0
## ✔ dials 1.2.0 ✔ tune 1.1.2
## ✔ infer 1.0.6 ✔ workflows 1.1.4
## ✔ modeldata 1.3.0 ✔ workflowsets 1.0.1
## ✔ parsnip 1.2.0 ✔ yardstick 1.2.0
## ✔ recipes 1.0.10
## ── Conflicts ───────────────────────────────────────── tidymodels_conflicts() ──
## ✖ scales::discard() masks purrr::discard()
## ✖ dplyr::filter() masks stats::filter()
## ✖ recipes::fixed() masks stringr::fixed()
## ✖ dplyr::lag() masks stats::lag()
## ✖ yardstick::spec() masks readr::spec()
## ✖ recipes::step() masks stats::step()
## • Use suppressPackageStartupMessages() to eliminate package startup messageslibrary(tidyquant)## Loading required package: lubridate
##
## Attaching package: 'lubridate'
##
## The following objects are masked from 'package:h2o':
##
## day, hour, month, week, year
##
## The following objects are masked from 'package:base':
##
## date, intersect, setdiff, union
##
## Loading required package: PerformanceAnalytics
## Loading required package: xts
## Loading required package: zoo
##
## Attaching package: 'zoo'
##
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric
##
##
## Attaching package: 'xts'
##
## The following objects are masked from 'package:dplyr':
##
## first, last
##
##
## Attaching package: 'PerformanceAnalytics'
##
## The following object is masked from 'package:graphics':
##
## legend
##
## Loading required package: quantmod
## Loading required package: TTR
##
## Attaching package: 'TTR'
##
## The following object is masked from 'package:dials':
##
## momentum
##
## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zoodata <- read_csv("../00_data/data_wrangled/data_Clean.csv") %>%
# h2o requires all variables to be either numeric or factors
mutate(across(where(is.character), factor))## Rows: 1470 Columns: 32
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (8): Attrition, BusinessTravel, Department, EducationField, Gender, Job...
## dbl (24): Age, DailyRate, DistanceFromHome, Education, EmployeeNumber, Envir...
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.set.seed(1234)
data_split <- initial_split(data, strata = "Attrition")
train_tbl <- training(data_split)
test_tbl <- testing(data_split)recipe_obj <- recipe(Attrition ~ ., data = train_tbl) %>%
# Remove zero variance variables
step_zv(all_predictors()) # Initialize h20
h2o.init()## Connection successful!
##
## R is connected to the H2O cluster:
## H2O cluster uptime: 4 days 18 hours
## H2O cluster timezone: America/New_York
## H2O data parsing timezone: UTC
## H2O cluster version: 3.44.0.3
## H2O cluster version age: 4 months and 9 days
## H2O cluster name: H2O_started_from_R_stephenmorris_fhp551
## H2O cluster total nodes: 1
## H2O cluster total memory: 0.93 GB
## H2O cluster total cores: 8
## H2O cluster allowed cores: 8
## H2O cluster healthy: TRUE
## H2O Connection ip: localhost
## H2O Connection port: 54321
## H2O Connection proxy: NA
## H2O Internal Security: FALSE
## R Version: R version 4.2.2 (2022-10-31)## Warning in h2o.clusterInfo():
## Your H2O cluster version is (4 months and 9 days) old. There may be a newer version available.
## Please download and install the latest version from: https://h2o-release.s3.amazonaws.com/h2o/latest_stable.htmlsplit.h2o <- h2o.splitFrame(as.h2o(train_tbl), ratios = c(0.85), seed = 2345)##
|
| | 0%
|
|======================================================================| 100%train_h2o <- split.h2o[[1]]
valid_h2o <- split.h2o[[2]]
test_h2o <- as.h2o(test_tbl)##
|
| | 0%
|
|======================================================================| 100%y <- "Attrition"
x <- setdiff(names(train_tbl), y)
models_h2o <- h2o.automl(
x = x,
y = y,
training_frame = train_h2o,
validation_frame = valid_h2o,
leaderboard_frame = test_h2o,
max_runtime_secs = 30,
nfolds = 5,
seed = 3456
)##
|
| | 0%
|
|== | 3%
## 11:02:18.692: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.
|
|======= | 10%
|
|============ | 17%
|
|================= | 24%
|
|===================== | 30%
|
|========================== | 37%
|
|=============================== | 44%
|
|=================================== | 51%
|
|======================================== | 57%
|
|============================================= | 64%
|
|================================================== | 71%
|
|====================================================== | 78%
|
|=========================================================== | 85%
|
|================================================================ | 91%
|
|===================================================================== | 98%
|
|======================================================================| 100%Examine the output of h2o.automl
models_h2o %>% typeof()## [1] "S4"models_h2o %>% slotNames()## [1] "project_name" "leader" "leaderboard" "event_log"
## [5] "modeling_steps" "training_info"models_h2o@leaderboard## model_id auc logloss
## 1 DeepLearning_grid_1_AutoML_13_20240430_110218_model_1 0.8648867 0.3968918
## 2 GBM_grid_1_AutoML_13_20240430_110218_model_1 0.8589536 0.3269150
## 3 StackedEnsemble_BestOfFamily_1_AutoML_13_20240430_110218 0.8558252 0.3076128
## 4 GBM_1_AutoML_13_20240430_110218 0.8552319 0.3243584
## 5 StackedEnsemble_BestOfFamily_4_AutoML_13_20240430_110218 0.8543689 0.3244397
## 6 XGBoost_grid_1_AutoML_13_20240430_110218_model_3 0.8510787 0.3242023
## aucpr mean_per_class_error rmse mse
## 1 0.6917862 0.2043689 0.3177864 0.10098819
## 2 0.6193157 0.2031553 0.3141580 0.09869523
## 3 0.6460828 0.2050162 0.3011894 0.09071507
## 4 0.6161812 0.2221683 0.3118296 0.09723772
## 5 0.6578135 0.1987864 0.2995784 0.08974721
## 6 0.5691088 0.2086570 0.3135505 0.09831391
##
## [50 rows x 7 columns]best_model <- models_h2o@leader
best_model## Model Details:
## ==============
##
## H2OBinomialModel: deeplearning
## Model ID: DeepLearning_grid_1_AutoML_13_20240430_110218_model_1
## Status of Neuron Layers: predicting Attrition, 2-class classification, bernoulli distribution, CrossEntropy loss, 6,202 weights/biases, 83.3 KB, 8,202 training samples, mini-batch size 1
## layer units type dropout l1 l2 mean_rate rate_rms
## 1 1 59 Input 5.00 % NA NA NA NA
## 2 2 100 RectifierDropout 20.00 % 0.000000 0.000000 0.118182 0.316937
## 3 3 2 Softmax NA 0.000000 0.000000 0.000417 0.000069
## momentum mean_weight weight_rms mean_bias bias_rms
## 1 NA NA NA NA NA
## 2 0.000000 0.000248 0.110865 0.488417 0.023293
## 3 0.000000 -0.043919 0.559701 -0.000824 0.012990
##
##
## H2OBinomialMetrics: deeplearning
## ** Reported on training data. **
## ** Metrics reported on full training frame **
##
## MSE: 0.09634606
## RMSE: 0.3103966
## LogLoss: 0.3816405
## Mean Per-Class Error: 0.2015651
## AUC: 0.8712267
## AUCPR: 0.6984956
## Gini: 0.7424535
##
## Confusion Matrix (vertical: actual; across: predicted) for F1-optimal threshold:
## No Yes Error Rate
## No 733 55 0.069797 =55/788
## Yes 50 100 0.333333 =50/150
## Totals 783 155 0.111940 =105/938
##
## Maximum Metrics: Maximum metrics at their respective thresholds
## metric threshold value idx
## 1 max f1 0.089085 0.655738 148
## 2 max f2 0.059804 0.683544 178
## 3 max f0point5 0.164437 0.716724 104
## 4 max accuracy 0.164437 0.902985 104
## 5 max precision 0.989795 1.000000 0
## 6 max recall 0.000013 1.000000 399
## 7 max specificity 0.989795 1.000000 0
## 8 max absolute_mcc 0.148798 0.606998 108
## 9 max min_per_class_accuracy 0.022760 0.780457 252
## 10 max mean_per_class_accuracy 0.063665 0.808443 172
## 11 max tns 0.989795 788.000000 0
## 12 max fns 0.989795 149.000000 0
## 13 max fps 0.000013 788.000000 399
## 14 max tps 0.000013 150.000000 399
## 15 max tnr 0.989795 1.000000 0
## 16 max fnr 0.989795 0.993333 0
## 17 max fpr 0.000013 1.000000 399
## 18 max tpr 0.000013 1.000000 399
##
## Gains/Lift Table: Extract with `h2o.gainsLift(<model>, <data>)` or `h2o.gainsLift(<model>, valid=<T/F>, xval=<T/F>)`
## H2OBinomialMetrics: deeplearning
## ** Reported on validation data. **
## ** Metrics reported on full validation frame **
##
## MSE: 0.1155056
## RMSE: 0.3398611
## LogLoss: 0.4427683
## Mean Per-Class Error: 0.2553105
## AUC: 0.8270697
## AUCPR: 0.5655039
## Gini: 0.6541394
##
## Confusion Matrix (vertical: actual; across: predicted) for F1-optimal threshold:
## No Yes Error Rate
## No 127 9 0.066176 =9/136
## Yes 12 15 0.444444 =12/27
## Totals 139 24 0.128834 =21/163
##
## Maximum Metrics: Maximum metrics at their respective thresholds
## metric threshold value idx
## 1 max f1 0.142432 0.588235 23
## 2 max f2 0.005541 0.634518 88
## 3 max f0point5 0.142432 0.609756 23
## 4 max accuracy 0.344682 0.871166 9
## 5 max precision 0.987339 1.000000 0
## 6 max recall 0.001059 1.000000 126
## 7 max specificity 0.987339 1.000000 0
## 8 max absolute_mcc 0.142432 0.513433 23
## 9 max min_per_class_accuracy 0.030671 0.705882 59
## 10 max mean_per_class_accuracy 0.058284 0.745098 41
## 11 max tns 0.987339 136.000000 0
## 12 max fns 0.987339 26.000000 0
## 13 max fps 0.000004 136.000000 162
## 14 max tps 0.001059 27.000000 126
## 15 max tnr 0.987339 1.000000 0
## 16 max fnr 0.987339 0.962963 0
## 17 max fpr 0.000004 1.000000 162
## 18 max tpr 0.001059 1.000000 126
##
## Gains/Lift Table: Extract with `h2o.gainsLift(<model>, <data>)` or `h2o.gainsLift(<model>, valid=<T/F>, xval=<T/F>)`
## H2OBinomialMetrics: deeplearning
## ** Reported on cross-validation data. **
## ** 5-fold cross-validation on training data (Metrics computed for combined holdout predictions) **
##
## MSE: 0.153224
## RMSE: 0.3914383
## LogLoss: 1.137627
## Mean Per-Class Error: 0.2999746
## AUC: 0.787923
## AUCPR: 0.504182
## Gini: 0.575846
##
## Confusion Matrix (vertical: actual; across: predicted) for F1-optimal threshold:
## No Yes Error Rate
## No 725 63 0.079949 =63/788
## Yes 78 72 0.520000 =78/150
## Totals 803 135 0.150320 =141/938
##
## Maximum Metrics: Maximum metrics at their respective thresholds
## metric threshold value idx
## 1 max f1 0.001260 0.505263 132
## 2 max f2 0.000124 0.604167 295
## 3 max f0point5 0.003373 0.547809 87
## 4 max accuracy 0.005941 0.864606 70
## 5 max precision 0.868487 1.000000 0
## 6 max recall 0.000000 1.000000 399
## 7 max specificity 0.868487 1.000000 0
## 8 max absolute_mcc 0.001604 0.419685 125
## 9 max min_per_class_accuracy 0.000154 0.720812 280
## 10 max mean_per_class_accuracy 0.000146 0.733266 285
## 11 max tns 0.868487 788.000000 0
## 12 max fns 0.868487 149.000000 0
## 13 max fps 0.000000 788.000000 399
## 14 max tps 0.000000 150.000000 399
## 15 max tnr 0.868487 1.000000 0
## 16 max fnr 0.868487 0.993333 0
## 17 max fpr 0.000000 1.000000 399
## 18 max tpr 0.000000 1.000000 399
##
## Gains/Lift Table: Extract with `h2o.gainsLift(<model>, <data>)` or `h2o.gainsLift(<model>, valid=<T/F>, xval=<T/F>)`
## Cross-Validation Metrics Summary:
## mean sd cv_1_valid cv_2_valid cv_3_valid
## accuracy 0.829326 0.085438 0.904255 0.808511 0.867021
## auc 0.761476 0.059887 0.802954 0.725527 0.745992
## err 0.170674 0.085438 0.095745 0.191489 0.132979
## err_count 32.000000 15.953056 18.000000 36.000000 25.000000
## f0point5 0.544804 0.162463 0.755814 0.447368 0.573770
## f1 0.521916 0.078442 0.590909 0.485714 0.528302
## f2 0.523196 0.037940 0.485075 0.531250 0.489510
## lift_top_group 5.630000 1.405070 6.266667 6.266667 6.266667
## logloss 1.242739 0.256118 0.802022 1.430429 1.258825
## max_per_class_error 0.460000 0.092496 0.566667 0.433333 0.533333
## mcc 0.446174 0.125207 0.595638 0.376800 0.457883
## mean_per_class_accuracy 0.712171 0.025538 0.713502 0.710549 0.704852
## mean_per_class_error 0.287829 0.025538 0.286498 0.289451 0.295148
## mse 0.155740 0.008159 0.141193 0.158987 0.158484
## pr_auc 0.512051 0.129236 0.649604 0.415577 0.545015
## precision 0.577203 0.240054 0.928571 0.425000 0.608696
## r2 -0.159224 0.059542 -0.052813 -0.185491 -0.181741
## recall 0.540000 0.092496 0.433333 0.566667 0.466667
## rmse 0.394527 0.010526 0.375757 0.398731 0.398100
## specificity 0.884343 0.117390 0.993671 0.854430 0.943038
## cv_4_valid cv_5_valid
## accuracy 0.689840 0.877005
## auc 0.692144 0.840764
## err 0.310160 0.122995
## err_count 58.000000 23.000000
## f0point5 0.331126 0.615942
## f1 0.408163 0.596491
## f2 0.531915 0.578231
## lift_top_group 3.116667 6.233333
## logloss 1.409152 1.313270
## max_per_class_error 0.333333 0.433333
## mcc 0.275366 0.525185
## mean_per_class_accuracy 0.680467 0.751486
## mean_per_class_error 0.319533 0.248514
## mse 0.159976 0.160060
## pr_auc 0.343212 0.606849
## precision 0.294118 0.629630
## r2 -0.187726 -0.188351
## recall 0.666667 0.566667
## rmse 0.399970 0.400075
## specificity 0.694268 0.936306?h2o.getModel
?h2o.saveModel
?h2o.loadModel
# h2o.getModel("DeepLearning_grid_1_AutoML_5_20240430_101059_model_1") %>%
# h2o.saveModel("h2o_models/")
#
# best_model <- h2o.loadModel("h2o_models/DeepLearning_grid_1_AutoML_5_20240430_101059_model_1")predictions <- h2o.predict(best_model, newdata = test_h2o)##
|
| | 0%
|
|======================================================================| 100%predictions_tbl <- predictions %>%
as_tibble()
predictions_tbl %>%
bind_cols(test_tbl)## # A tibble: 369 × 35
## predict No Yes Age Attrition BusinessTravel DailyRate Department
## <fct> <dbl> <dbl> <dbl> <fct> <fct> <dbl> <fct>
## 1 No 0.999 0.000581 59 No Travel_Rarely 1324 Research …
## 2 No 0.993 0.00674 35 No Travel_Rarely 809 Research …
## 3 No 0.995 0.00515 34 No Travel_Rarely 1346 Research …
## 4 No 0.986 0.0139 22 No Non-Travel 1123 Research …
## 5 No 1.00 0.0000325 53 No Travel_Rarely 1219 Sales
## 6 No 0.990 0.0103 24 No Non-Travel 673 Research …
## 7 No 0.979 0.0213 21 No Travel_Rarely 391 Research …
## 8 No 0.968 0.0317 34 Yes Travel_Rarely 699 Research …
## 9 No 1.00 0.000109 53 No Travel_Rarely 1282 Research …
## 10 Yes 0.190 0.810 32 Yes Travel_Frequent… 1125 Research …
## # ℹ 359 more rows
## # ℹ 27 more variables: DistanceFromHome <dbl>, Education <dbl>,
## # EducationField <fct>, EmployeeNumber <dbl>, EnvironmentSatisfaction <dbl>,
## # Gender <fct>, HourlyRate <dbl>, JobInvolvement <dbl>, JobLevel <dbl>,
## # JobRole <fct>, JobSatisfaction <dbl>, MaritalStatus <fct>,
## # MonthlyIncome <dbl>, MonthlyRate <dbl>, NumCompaniesWorked <dbl>,
## # OverTime <fct>, PercentSalaryHike <dbl>, PerformanceRating <dbl>, …?h2o.performance
performance_h2o <- h2o.performance(best_model, newdata = test_h2o)
typeof(performance_h2o)## [1] "S4"slotNames(performance_h2o)## [1] "algorithm" "on_train" "on_valid" "on_xval" "metrics"performance_h2o@metrics## $model
## $model$`__meta`
## $model$`__meta`$schema_version
## [1] 3
##
## $model$`__meta`$schema_name
## [1] "ModelKeyV3"
##
## $model$`__meta`$schema_type
## [1] "Key<Model>"
##
##
## $model$name
## [1] "DeepLearning_grid_1_AutoML_13_20240430_110218_model_1"
##
## $model$type
## [1] "Key<Model>"
##
## $model$URL
## [1] "/3/Models/DeepLearning_grid_1_AutoML_13_20240430_110218_model_1"
##
##
## $model_checksum
## [1] "-7290620826851224128"
##
## $frame
## $frame$name
## [1] "test_tbl_sid_9b97_3"
##
##
## $frame_checksum
## [1] "-54413681510283746"
##
## $description
## NULL
##
## $scoring_time
## [1] 1.714489e+12
##
## $predictions
## NULL
##
## $MSE
## [1] 0.1009882
##
## $RMSE
## [1] 0.3177864
##
## $nobs
## [1] 369
##
## $custom_metric_name
## NULL
##
## $custom_metric_value
## [1] 0
##
## $r2
## [1] 0.2583251
##
## $logloss
## [1] 0.3968918
##
## $AUC
## [1] 0.8648867
##
## $pr_auc
## [1] 0.6917862
##
## $Gini
## [1] 0.7297735
##
## $mean_per_class_error
## [1] 0.2043689
##
## $domain
## [1] "No" "Yes"
##
## $cm
## $cm$`__meta`
## $cm$`__meta`$schema_version
## [1] 3
##
## $cm$`__meta`$schema_name
## [1] "ConfusionMatrixV3"
##
## $cm$`__meta`$schema_type
## [1] "ConfusionMatrix"
##
##
## $cm$table
## Confusion Matrix: Row labels: Actual class; Column labels: Predicted class
## No Yes Error Rate
## No 296 13 0.0421 = 13 / 309
## Yes 22 38 0.3667 = 22 / 60
## Totals 318 51 0.0949 = 35 / 369
##
##
## $thresholds_and_metric_scores
## Metrics for Thresholds: Binomial metrics as a function of classification thresholds
## threshold f1 f2 f0point5 accuracy precision recall specificity
## 1 0.976659 0.032787 0.020747 0.078125 0.840108 1.000000 0.016667 1.000000
## 2 0.929877 0.064516 0.041322 0.147059 0.842818 1.000000 0.033333 1.000000
## 3 0.875176 0.095238 0.061728 0.208333 0.845528 1.000000 0.050000 1.000000
## 4 0.845200 0.125000 0.081967 0.263158 0.848238 1.000000 0.066667 1.000000
## 5 0.844310 0.153846 0.102041 0.312500 0.850949 1.000000 0.083333 1.000000
## absolute_mcc min_per_class_accuracy mean_per_class_accuracy tns fns fps tps
## 1 0.118299 0.016667 0.508333 309 59 0 1
## 2 0.167527 0.033333 0.516667 309 58 0 2
## 3 0.205458 0.050000 0.525000 309 57 0 3
## 4 0.237568 0.066667 0.533333 309 56 0 4
## 5 0.265973 0.083333 0.541667 309 55 0 5
## tnr fnr fpr tpr idx
## 1 1.000000 0.983333 0.000000 0.016667 0
## 2 1.000000 0.966667 0.000000 0.033333 1
## 3 1.000000 0.950000 0.000000 0.050000 2
## 4 1.000000 0.933333 0.000000 0.066667 3
## 5 1.000000 0.916667 0.000000 0.083333 4
##
## ---
## threshold f1 f2 f0point5 accuracy precision recall
## 364 0.000008 0.283019 0.496689 0.197889 0.176152 0.164835 1.000000
## 365 0.000005 0.282353 0.495868 0.197368 0.173442 0.164384 1.000000
## 366 0.000004 0.281690 0.495050 0.196850 0.170732 0.163934 1.000000
## 367 0.000003 0.281030 0.494234 0.196335 0.168022 0.163488 1.000000
## 368 0.000001 0.280374 0.493421 0.195822 0.165312 0.163043 1.000000
## 369 0.000000 0.279720 0.492611 0.195313 0.162602 0.162602 1.000000
## specificity absolute_mcc min_per_class_accuracy mean_per_class_accuracy tns
## 364 0.016181 0.051645 0.016181 0.508091 5
## 365 0.012945 0.046130 0.012945 0.506472 4
## 366 0.009709 0.039895 0.009709 0.504854 3
## 367 0.006472 0.032530 0.006472 0.503236 2
## 368 0.003236 0.022971 0.003236 0.501618 1
## 369 0.000000 0.000000 0.000000 0.500000 0
## fns fps tps tnr fnr fpr tpr idx
## 364 0 304 60 0.016181 0.000000 0.983819 1.000000 363
## 365 0 305 60 0.012945 0.000000 0.987055 1.000000 364
## 366 0 306 60 0.009709 0.000000 0.990291 1.000000 365
## 367 0 307 60 0.006472 0.000000 0.993528 1.000000 366
## 368 0 308 60 0.003236 0.000000 0.996764 1.000000 367
## 369 0 309 60 0.000000 0.000000 1.000000 1.000000 368
##
## $max_criteria_and_metric_scores
## Maximum Metrics: Maximum metrics at their respective thresholds
## metric threshold value idx
## 1 max f1 0.113174 0.684685 50
## 2 max f2 0.014028 0.670103 147
## 3 max f0point5 0.178196 0.727273 39
## 4 max accuracy 0.113174 0.905149 50
## 5 max precision 0.976659 1.000000 0
## 6 max recall 0.000199 1.000000 332
## 7 max specificity 0.976659 1.000000 0
## 8 max absolute_mcc 0.113174 0.632173 50
## 9 max min_per_class_accuracy 0.025512 0.783333 111
## 10 max mean_per_class_accuracy 0.113174 0.795631 50
## 11 max tns 0.976659 309.000000 0
## 12 max fns 0.976659 59.000000 0
## 13 max fps 0.000000 309.000000 368
## 14 max tps 0.000199 60.000000 332
## 15 max tnr 0.976659 1.000000 0
## 16 max fnr 0.976659 0.983333 0
## 17 max fpr 0.000000 1.000000 368
## 18 max tpr 0.000199 1.000000 332
##
## $gains_lift_table
## Gains/Lift Table: Avg response rate: 16.26 %, avg score: 6.73 %
## group cumulative_data_fraction lower_threshold lift cumulative_lift
## 1 1 0.01084011 0.844595 6.150000 6.150000
## 2 2 0.02168022 0.726109 6.150000 6.150000
## 3 3 0.03252033 0.578426 6.150000 6.150000
## 4 4 0.04065041 0.479439 2.050000 5.330000
## 5 5 0.05149051 0.381714 4.612500 5.178947
## 6 6 0.10027100 0.206685 4.441667 4.820270
## 7 7 0.15176152 0.103748 2.913158 4.173214
## 8 8 0.20054201 0.065526 0.683333 3.324324
## 9 9 0.30081301 0.026618 0.997297 2.548649
## 10 10 0.40108401 0.014017 0.997297 2.160811
## 11 11 0.50135501 0.007349 0.332432 1.795135
## 12 12 0.59891599 0.004044 0.341667 1.558371
## 13 13 0.69918699 0.002067 0.166216 1.358721
## 14 14 0.79945799 0.000646 0.166216 1.209153
## 15 15 0.89972900 0.000212 0.166216 1.092922
## 16 16 1.00000000 0.000000 0.166216 1.000000
## response_rate score cumulative_response_rate cumulative_score
## 1 1.000000 0.906728 1.000000 0.906728
## 2 1.000000 0.803167 1.000000 0.854948
## 3 1.000000 0.618807 1.000000 0.776234
## 4 0.333333 0.519289 0.866667 0.724845
## 5 0.750000 0.428765 0.842105 0.662512
## 6 0.722222 0.283616 0.783784 0.478184
## 7 0.473684 0.139358 0.678571 0.363226
## 8 0.111111 0.082208 0.540541 0.294870
## 9 0.162162 0.042455 0.414414 0.210732
## 10 0.162162 0.018852 0.351351 0.162762
## 11 0.054054 0.010349 0.291892 0.132279
## 12 0.055556 0.005652 0.253394 0.111652
## 13 0.027027 0.002863 0.220930 0.096050
## 14 0.027027 0.001240 0.196610 0.084159
## 15 0.027027 0.000403 0.177711 0.074825
## 16 0.027027 0.000075 0.162602 0.067329
## capture_rate cumulative_capture_rate gain cumulative_gain
## 1 0.066667 0.066667 515.000000 515.000000
## 2 0.066667 0.133333 515.000000 515.000000
## 3 0.066667 0.200000 515.000000 515.000000
## 4 0.016667 0.216667 105.000000 433.000000
## 5 0.050000 0.266667 361.250000 417.894737
## 6 0.216667 0.483333 344.166667 382.027027
## 7 0.150000 0.633333 191.315789 317.321429
## 8 0.033333 0.666667 -31.666667 232.432432
## 9 0.100000 0.766667 -0.270270 154.864865
## 10 0.100000 0.866667 -0.270270 116.081081
## 11 0.033333 0.900000 -66.756757 79.513514
## 12 0.033333 0.933333 -65.833333 55.837104
## 13 0.016667 0.950000 -83.378378 35.872093
## 14 0.016667 0.966667 -83.378378 20.915254
## 15 0.016667 0.983333 -83.378378 9.292169
## 16 0.016667 1.000000 -83.378378 0.000000
## kolmogorov_smirnov
## 1 0.066667
## 2 0.133333
## 3 0.200000
## 4 0.210194
## 5 0.256958
## 6 0.457443
## 7 0.575081
## 8 0.556634
## 9 0.556311
## 10 0.555987
## 11 0.476052
## 12 0.399353
## 13 0.299515
## 14 0.199676
## 15 0.099838
## 16 0.000000h2o.auc(performance_h2o)## [1] 0.8648867h2o.confusionMatrix(performance_h2o)## Confusion Matrix (vertical: actual; across: predicted) for max f1 @ threshold = 0.113174371154645:
## No Yes Error Rate
## No 296 13 0.042071 =13/309
## Yes 22 38 0.366667 =22/60
## Totals 318 51 0.094851 =35/369