# On the poor performance of classifiers in insurance models

Each time we have a case study in my actuarial courses (with real data), students are surprised to have hard time getting a “good” model, and they are always surprised to have a low AUC, when trying to model the probability to claim a loss, to die, to fraud, etc. And each time, I keep saying, “yes, I know, and that’s what we expect because there a lot of ‘randomness’ in insurance”. To be more specific, I decided to run some simulations, and to compute AUCs to see what’s going on. And because I don’t want to waste time fitting models, we will assume that we have each time a perfect model. So I want to show that the upper bound of the AUC is actually quite low ! So it’s not a modeling issue, it is a fondamental issue in insurance !

By ‘perfect model’ I mean the following : $\Omega$ denotes the heterogeneity factor, because people are different. We would love to get $\mathbb{P}[Y=1|\Omega]$. Unfortunately, $\Omega$  is unobservable ! So we use covariates (like the age of the driver of the car in motor insurance, or of the policyholder in life insurance, etc). Thus, we have data $(y_i,\boldsymbol{x}_i)$‘s and we use them to train a model, in order to approximate $\mathbb{P}[Y=1|\boldsymbol{X}]$. And then, we check if our model is good (or not) using the ROC curve, obtained from confusion matrices, comparing $y_i$‘s and $\widehat{y}_i$‘s where $\widehat{y}_i=1$ when $\mathbb{P}[Y_i=1|\boldsymbol{x}_i]$ exceeds a given threshold. Here, I will not try to construct models. I will predict $\widehat{y}_i=1$ each time the true underlying probability $\mathbb{P}[Y_i=1|\omega_i]$ exceeds a threshold ! The point is that it’s possible to claim a loss ($y=1$) even if the probability is 3% (and most of the time $\widehat{y}=0$), and to not claim one ($y=0$) even if the probability is 97% (and most of the time $\widehat{y}=1$). That’s the idea with randomness, right ?

So, here $p(\omega_1),\cdots,p(\omega_n)$ denote the probabilities to claim a loss, to die, to fraud, etc. There is heterogeneity here, and this heterogenity can be small, or large. Consider the graph below, to illustrate, In both cases, there is, on average, 25% chance to claim a loss. But on the left, there is more heterogeneity, more dispersion. To illustrate, I used the arrow, which is a classical 90% interval : 90% of the individuals have a probability to claim a loss in that interval. (here 10%-40%), 5% are below 10% (low risk), and 5% are above 40% (high risk). Later on, we will say that we have 25% on average, with a dispersion of 30% (40% minus 10%). On the right, it’s more 25% on average, with a dispersion of of 15%. What I call dispersion is the difference between the 95% and the 5% quantiles.

Consider now some dataset, with Bernoulli variables $y$, drawn with those probabilities $p(\omega)$. Then, let us assume that we are able to get a perfect model : I do not estimate a model based on some covariates, here, I assume that I know perfectly the probability (which is true, because I did generate those data). More specifically, to generate a vector of probabilities, here I use a Beta distribution with a given mean, and a given variance (to capture the heterogeneity I mentioned above)

a=m*(m*(1-m)/v-1) b=(1-m)*(m*(1-m)/v-1) p=rbeta(n,a,b)

from those probabilities, I generate occurences of claims, or deaths,

Y=rbinom(n,size = 1,prob = p)

Then, I compute the AUC of my “perfect” model,

auc.tmp=performance(prediction(p,Y),"auc")

And then, I will generate many samples, to compute the average value of the AUC. And actually, we can do that for many values of the mean and the variance of the Beta distribution. Here is the code

library(ROCR) n=1000 ns=200 ab_beta = function(m,inter){ a=uniroot(function(a) qbeta(.95,a,a/m-a)-qbeta(.05,a,a/m-a)-inter, interval=c(.0000001,1000000))$root b=a/m-a return(c(a,b)) } Sim_AUC_mean_inter=function(m=.5,i=.05){ V_auc=rep(NA,ns) b=-1 essai = try(ab&lt;-ab_beta(m,i),TRUE) if(inherits(essai,what="try-error")) a=-1 if(!inherits(essai,what="try-error")){ a=ab b=ab } if((a&gt;=0)&amp;(b&gt;=0)){ for(s in 1:ns){ p=rbeta(n,a,b) Y=rbinom(n,size = 1,prob = p) auc.tmp=performance(prediction(p,Y),"auc") V_auc[s]=as.numeric(auc.tmp@y.values)} L=list(moy_beta=m, var_beat=v, q05=qbeta(.05,a,b), q95=qbeta(.95,a,b), moy_AUC=mean(V_auc), sd_AUC=sd(V_auc), q05_AUC=quantile(V_auc,.05), q95_AUC=quantile(V_auc,.95)) return(L)} if((a&lt;0)|(b&lt;0)){return(list(moy_AUC=NA))}} Vm=seq(.025,.975,by=.025) Vi=seq(.01,.5,by=.01) V=outer(X = Vm,Y = Vi, Vectorize(function(x,y) Sim_AUC_mean_inter(x,y)$moy_AUC)) library("RColorBrewer") image(Vm,Vi,V, xlab="Probability (Average)", ylab="Dispersion (Q95-Q5)", col= colorRampPalette(brewer.pal(n = 9, name = "YlGn"))(101)) contour(Vm,Vi,V,add=TRUE,lwd=2) On the x-axis, we have the average probability to claim a loss. Of course, there is a symmetry here. And on the y-axis, we have the dispersion : the lower, the less heterogeneity in the portfolio. For instance, with a 30% chance to claim a loss on average, and 20% dispersion (meaning that in the portfolio, 90% of the insured have between 20% and 40% chance to claim a loss, or 15% and 35% chance), we have on average a 60% AUC. With a perfect model ! So with only a few covariates, having 55% should be great !

My point here is that with a low dispersion, we cannot expect to have a great AUC (again, even with a perfect model). In motor insurance, from my experience, 90% of the insured are between 3% chance and 20% chance to claim a loss ! That’s less than 20% dispersion ! and in that case, even if the (average) probability is rather small, it is very difficult to expect an AUC above 60% or 65% !

# Ce que la courbe ROC (et l’AUC) ne raconte pas

En préparant une intervention pour mardi prochain, j’épluchais les résultats renvoyés pour un exercice, et j’ai eu un résultat assez étrange avec un modèle de classification. J’avais donné la même base cet automne à l’ensae, et j’avais donc près d’une trentaine d’autres modèles, pour comparer (disons plutôt que sur la même base de test, j’ai une trentaine de prévisions). Les observations noires sont celles obtenues cet automne (le trait correspond aux meilleurs AUC sur la base de test), et les observations rouges sont celles que j’ai obtenu pour l’intervention de mardi (là encore, le trait vertical correspond aux meilleurs modèles, au sens du AUC), sur une observation de la base de test, Ce sont les probabilités prédites (mais j’ai enlevé l’échelle).

Pour presque toutes mes observations, les poids rouges sont bien au dessus des autres… Mais cela n’enlève en rien le fait que l’AUC obtenu (pour les deux modèles rouges) est très bon. C’est effectivement un résultat important (et connu) : le critère AUC (mais plus généralement la courbe ROC en entier) n’indique en rien si la valeur prédite est bonne, ou pas. Il nous dit juste si l’ordre obtenu est correct. Si les valeurs les plus importantes sont effectivement les valeurs pour lesquelles on a un 1, l’AUC sera très bon.

C’est ce qu’on peut observer sur le petit exemple ci-dessous. Considérons un modèle logistique simulé assez simple,

> n=1e3
> set.seed(1)
> x1=rnorm(n)
> x2=runif(n)
> u=-3+x2+x1
> p=exp(u)/(1+exp(u))
> y=rbinom(n,prob=p,size=1)
> library(ROCR)
> df=data.frame(y,x1,x2)
> mean(df$y)  0.116 > reg=glm(y~.,data=df,family=binomial) > p=predict(reg,type="response") > mean(p)  0.116 > pred1=prediction(p, df$y)
> L=performance(pred1, "tpr", "fpr")

L’AUC est ici

> auc=performance(pred1, "auc")@y.values[]
> auc
 0.7681191

et la courbe ROC est la suivante,

> plot(unlist(L@x.values),unlist(L@y.values),
+ type="s",col="blue") Supposons que l’on change la constante du modèle logistique,

> reg$coefficients=0 Dans ce cas, notre prévision est assez mauvaise, car la probabilité moyenne prédite est ici > u=reg$coefficients+reg$coefficients* + df$x1+reg$coefficients*df$x2
> p=exp(u)/(1+exp(u))
> mean(p)
 0.6060676

(on est loin des 11,6% de 1 dans la base). Pourtant le AUC est bon

> pred1=prediction(p, df$y) > L=performance(pred1, "tpr", "fpr") > auc=performance(pred1, "auc")@y.values[] > auc  0.7681191 (c’est en fait la même valeur qu’auparavant, ce qui a du sens puisque la courbe ROC est identique) > lines(unlist(L@x.values),unlist(L@y.values), + type="s",col="red") Autrement dit, ces outils, classiquement utilisés pour juger la qualité d’un classifieur, ne permettent en aucun cas de dire que la probabilité prédite a du sens. Ces critères permettent juste de dire qu’on identifie assez bien les personnes qui ont le plus de chance d’avoir la réponse 1. Ce qui n’est pas si mal… mais c’est un autre problème que celui d’avoir une probabilité qui soit pertinente. # Classification on the German Credit Database In our data science course, this morning, we’ve use random forrest to improve prediction on the German Credit Dataset. The dataset is > url="http://freakonometrics.free.fr/german_credit.csv" > credit=read.csv(url, header = TRUE, sep = ",") Almost all variables are treated a numeric, but actually, most of them are factors, > str(credit) 'data.frame': 1000 obs. of 21 variables:$ Creditability   : int  1 1 1 1 1 1 1 1 1 1 ...
$Account.Balance : int 1 1 2 1 1 1 1 1 4 2 ...$ Duration        : int  18 9 12 12 12 10 8  ...
$Purpose : int 2 0 9 0 0 0 0 0 3 3 ... (etc). Let us convert categorical variables as factors, > F=c(1,2,4,5,7,8,9,10,11,12,13,15,16,17,18,19,20) > for(i in F) credit[,i]=as.factor(credit[,i]) Let us now create our training/calibration and validation/testing datasets, with proportion 1/3-2/3 > i_test=sample(1:nrow(credit),size=333) > i_calibration=(1:nrow(credit))[-i_test] The first model we can fit is a logistic regression, on selected covariates > LogisticModel <- glm(Creditability ~ Account.Balance + Payment.Status.of.Previous.Credit + Purpose + Length.of.current.employment + Sex...Marital.Status, family=binomial, data = credit[i_calibration,]) Based on that model, it is possible to draw the ROC curve, and to compute the AUC (on ne validation dataset) > fitLog <- predict(LogisticModel,type="response", + newdata=credit[i_test,]) > library(ROCR) > pred = prediction( fitLog, credit$Creditability[i_test])
> perf <- performance(pred, "tpr", "fpr")
> plot(perf)
> AUCLog1=performance(pred, measure = "auc")@y.values[]
> cat("AUC: ",AUCLog1,"\n")
AUC:  0.7340997 An alternative is to consider a logistic regression on all explanatory variables

> LogisticModel <- glm(Creditability ~ .,
+  family=binomial,
+  data = credit[i_calibration,])

We might overfit, here, and we should observe that on the ROC curve

> fitLog <- predict(LogisticModel,type="response",
+                   newdata=credit[i_test,])
> pred = prediction( fitLog, credit$Creditability[i_test]) > perf <- performance(pred, "tpr", "fpr") > plot(perf) > AUCLog2=performance(pred, measure = "auc")@y.values[] > cat("AUC: ",AUCLog2,"\n") AUC: 0.7609792 There is a slight improvement here, compared with the previous model, where only five explanatory variables were considered. Consider now some regression tree (on all covariates) > library(rpart) > ArbreModel <- rpart(Creditability ~ ., + data = credit[i_calibration,]) We can visualize the tree using > library(rpart.plot) > prp(ArbreModel,type=2,extra=1) The ROC curve for that model is > fitArbre <- predict(ArbreModel, + newdata=credit[i_test,], + type="prob")[,2] > pred = prediction( fitArbre, credit$Creditability[i_test])
> perf <- performance(pred, "tpr", "fpr")
> plot(perf)
> AUCArbre=performance(pred, measure = "auc")@y.values[]
> cat("AUC: ",AUCArbre,"\n")
AUC:  0.7100323 As expected, a single has a lower performance, compared with a logistic regression. And a natural idea is to grow several trees using some boostrap procedure, and then to agregate those predictions.

> library(randomForest)
> RF <- randomForest(Creditability ~ .,
+ data = credit[i_calibration,])
> fitForet <- predict(RF,
+                     newdata=credit[i_test,],
+                     type="prob")[,2]
> pred = prediction( fitForet, credit$Creditability[i_test]) > perf <- performance(pred, "tpr", "fpr") > plot(perf) > AUCRF=performance(pred, measure = "auc")@y.values[] > cat("AUC: ",AUCRF,"\n") AUC: 0.7682367 Here this model is (slightly) better than the logistic regression. Actually, if we create many training/validation samples, and compare the AUC, we can observe that – on average – random forests perform better than logistic regressions, > AUC=function(i){ + set.seed(i) + i_test=sample(1:nrow(credit),size=333) + i_calibration=(1:nrow(credit))[-i_test] + LogisticModel <- glm(Creditability ~ ., + family=binomial, + data = credit[i_calibration,]) + summary(LogisticModel) + fitLog <- predict(LogisticModel,type="response", + newdata=credit[i_test,]) + library(ROCR) + pred = prediction( fitLog, credit$Creditability[i_test])
+   AUCLog2=performance(pred, measure = "auc")@y.values[]
+   RF <- randomForest(Creditability ~ .,
+   data = credit[i_calibration,])
+   fitForet <- predict(RF,
+                       newdata=credit[i_test,],
+                       type="prob")[,2]
+   pred = prediction( fitForet, credit\$Creditability[i_test])
+   AUCRF=performance(pred, measure = "auc")@y.values[]
+   return(c(AUCLog2,AUCRF))
+ }
> A=Vectorize(AUC)(1:200)
> plot(t(A)) 