Individual Driver Crash Risk Classification Based on IoV Data and Offline Consumer Behavior Data
Table 10
Comparison of consumer behavior-based models for severe collision and general collision risk classification (binning data).
Dependent variable
LRbox
NNbox
RFbox
SVMbox
AUC
IR (%)
AUC
IR (%)
AUC
IR (%)
AUC
IR (%)
Basic features
0.800
—
0.729
—
0.785
—
0.778
—
Basic features + proportion of maintenance carried out in 4S dealer shops
0.801
0.07
0.739
1.40
0.779
-0.75
0.783
0.64
Basic features + average price of maintenance carried out in 4S dealer shops
0.804
0.44
0.733
0.52
0.781
-0.61
0.785
0.89
Basic features + average maintenance interval (days)
0.855
6.91
0.820
12.47
0.824
4.95
0.829
6.57
Basic features + average maintenance interval (miles)
0.824
3.00
0.764
4.84
0.819
4.32
0.796
2.35
Basic features + average number of products used in one maintenance episode
0.801
0.07
0.731
0.25
0.789
0.42
0.780
0.24
Note 1: LRbox, NNbox, RFbox, and SVMbox represent LR, NN, RF, and SVM models using binning data, respectively. Note 2: IR represents the increase in the AUC compared with the basic feature-based model.