Bayesian Logistic Regression Models for Credit Scoring by Gregg Webster

Fig. 4.8
Histogram and Box plot of CLNO…………………………………………………..55 
Fig. 4.9
Histogram and Box plot of DELINQ………………………………………………..56 
Fig. 4.10
Histogram and Box plot of CLAGE……………………………………………….56 
Fig. 4.11
Histogram and Box plot of NINQ…………………………………………………56 
Fig. 4.12
Half-normal plot of residuals for the “old” data…………………………………...61 
Fig. 4.13
Half-normal plot of leverages for the “old” data…………………………………..62 
Fig. 4.14
Half-normal plot of the Cook’s distance statistics for the “old” data……………..62 
Fig. 4.15
Optimal cut-off probability when total error is minimized………………………...65 
Fig. 4.16 
Optimal cut-off probability when error function is minimized with α = 0.8………65 
Fig. 4.17
Half-normal plot of residuals for the “new” data………………………………….70 
Fig. 4.18
Half-normal plot of leverages for the “new” data………………………………….70 
Fig. 4.19
Half-normal plot of the Cook’s distance statistics for the “new” data…………….71 
Fig. 4.20
Trace and density plots of the posteriors for the first four variables using an 
informative prior…………………………………………………………………………..77 

---

9 
Fig. 4.21
Trace and density plots of the posteriors for the first four variables using a non-
informative prior…………………………………………………………………………..80 
Fig. 4.22
Error rates of Models 1, 2 and 3 when the models are trained using differerent 
sampler sizes and the cut-off probability is 0.3…………………………………………...88 
Fig. 4.23
Error rates of Models 1, 2 and 3 when the models are trained using differerent 
sampler sizes and the cut-off probability is 0.48……………………………………….....90 
 
 

---

10