Use your selected model to predict the probability of crossing
the hurdle for someone having the median values of all variables
(for any binary factors, determine which category is more
prevalent). Then, assuming they do have negative marks
against their credit, predict the number of such marks.
Use Your Selected Model To Predict The Probability Of Crossing The Hurdle For Someone Having The Median Values Of All Va 1 (74.24 KiB) Viewed 21 times
Call: hurdle(formula = reports - share + owner + active share + owner + months + active, data = newcc, dist = "negbin") = Pearson residuals: Min 1Q Median Max -0.8429 -0.3900 -0.3218 -0.2317 22.0073 30 Count model coefficients (truncated negbin with log link): Estimate Std. Error z value Pr(>Izı) (Intercept) -0.07899 0.35073 -0.225 0.82180 share -12.48409 2. 19171 -5.696 1.23e-08 *** owneryes -0.54936 0.20551 -2.673 0.00751 ** active 0.05781 0.01800 3.211 0.00132 ** Log(theta) -0.49534 0.49895 -0.993 0.32082 Zero hurdle model coefficients (binomial with logit link): Estimate Std. Error z value Pr(>Izl) (Intercept) -1.819685 0.143711 -12.662 <2e-16 *** share -5.394686 1.188025 -4.541 5.60e-06 *** owneryes -0.627525 0.161123 -3.895 9.83e-05 *** months 0.003758 0.001013 3.709 0.000208 *** active 0.096218 0.011586 8.304 < 2e-16 *** ! Signif. codes: 0 '***' 0.001 "**' 0.01 '*' 0.05 '.' 0.1 1 = Theta: count = 0.6094 Number of iterations in BFGS optimization: 21 Log-likelihood: -968.2 on 10 Df
Join a community of subject matter experts. Register for FREE to view solutions, replies, and use search function. Request answer by replying!