Car Wreck Predictions and Accuracy

Table of contents

• Setup
  • Data Frame Structure
• Final Predictions, Reality, and Accuracy (PLUS OTHER IMPORTANT STATS)

Setup

Data Frame Structure

'data.frame':   6529 obs. of  26 variables:
 $ V1 : chr  "INDEX" "8023" "4658" "2005" ...
 $ V2 : chr  "TARGET_FLAG" "0" "0" "0" ...
 $ V3 : chr  "TARGET_AMT" "0" "0" "0" ...
 $ V4 : chr  "KIDSDRIV" "0" "0" "0" ...
 $ V5 : chr  "AGE" "40" "48" "51" ...
 $ V6 : chr  "HOMEKIDS" "0" "0" "0" ...
 $ V7 : chr  "YOJ" "9" "0" "14" ...
 $ V8 : chr  "INCOME" "$78,658" "$0" "$89,606" ...
 $ V9 : chr  "PARENT1" "No" "No" "No" ...
 $ V10: chr  "HOME_VAL" "$236,005" "$75,688" "$288,266" ...
 $ V11: chr  "MSTATUS" "z_No" "Yes" "z_No" ...
 $ V12: chr  "SEX" "z_F" "z_F" "z_F" ...
 $ V13: chr  "EDUCATION" "Masters" "Masters" "Bachelors" ...
 $ V14: chr  "JOB" "Lawyer" "Home Maker" "Clerical" ...
 $ V15: chr  "TRAVTIME" "69" "39" "32" ...
 $ V16: chr  "CAR_USE" "Private" "Private" "Private" ...
 $ V17: chr  "BLUEBOOK" "$17,040" "$1,500" "$10,460" ...
 $ V18: chr  "TIF" "4" "6" "1" ...
 $ V19: chr  "CAR_TYPE" "Pickup" "Sports Car" "z_SUV" ...
 $ V20: chr  "RED_CAR" "no" "no" "no" ...
 $ V21: chr  "OLDCLAIM" "$0" "$3,867" "$21,581" ...
 $ V22: chr  "CLM_FREQ" "0" "3" "2" ...
 $ V23: chr  "REVOKED" "No" "No" "Yes" ...
 $ V24: chr  "MVR_PTS" "1" "5" "0" ...
 $ V25: chr  "CAR_AGE" "15" NA "8" ...
 $ V26: chr  "URBANICITY" "z_Highly Rural/ Rural" "Highly Urban/ Urban" "Highly Urban/ Urban" ...

Final Predictions, Reality, and Accuracy (PLUS OTHER IMPORTANT STATS)

Confusion Matrix and Statistics

          Reference
Prediction   0   1
         0 780 161
         1 355 244
                                          
               Accuracy : 0.6649          
                 95% CI : (0.6407, 0.6885)
    No Information Rate : 0.737           
    P-Value [Acc > NIR] : 1               
                                          
                  Kappa : 0.251           
                                          
 Mcnemar's Test P-Value : <2e-16          
                                          
            Sensitivity : 0.6025          
            Specificity : 0.6872          
         Pos Pred Value : 0.4073          
         Neg Pred Value : 0.8289          
             Prevalence : 0.2630          
         Detection Rate : 0.1584          
   Detection Prevalence : 0.3890          
      Balanced Accuracy : 0.6448          
                                          
       'Positive' Class : 1