Automobile insurance claim occurrence prediction model based on ensemble learning

Abstract

The generalized linear model (GLM) is a widely used method in traditional automobile insurance loss prediction. Ensemble learning algorithms have recently shown promising results in the realm of automobile insurance, providing a new option for loss prediction. In the age of big data, how to predict loss in automobile insurance more accurately is an urgent problem to be solved. Stacking is a hot issue in ensemble learning that has been effectively used in many fields, but few researchers have applied it to the field of automobile insurance. In this research, Stacking was innovatively introduced into loss prediction for automobile insurance to solve this problem. Three datasets related to automobile insurance were used. Adopting the Synthetic Minority Oversampling Technique for class balance, a model of claim occurrence was established using four methods, namely logistic regression in GLM and bagging, boosting, stacking in ensemble learning.The area under the receiver operating characteristic curve values and F1-scores achieved with the four methods were then compared to assess classification performance. Ensemble algorithms were used to rank the importance of features in the FRE dataset. Finally, we combined probability with the bonus-malus system to formulate a fairer transfer strategy. The results showed that the proposed approach performed better than the other methods on all datasets, with significantly enhanced prediction accuracy.

Prediction of Auto Insurance Risk Based on t-SNE Dimensionality Reduction

“…While the neural network gives the best performance, the other non-linear models are still considerably close to it. Although in general the values of the AUC presented here are quite low, these are very representative of the results in the field of auto insurance [21,22]. To further establish the robustness of our approach we compare the results in TABLE 1 to the same set of models without the stage of dimensional reduction, i.e., the models were trained directly on the high dimensional space (R 14 → [0,1]).…”

Freight insurance pricing strategy based on an online freight platform

“…A generalised linear model, which allows deviation from the normality assumption and extends the regression to the exponential distribution family, was first applied to auto insurance ratemaking by McCullagh and Nelder (1989). Research on the pricing of generalised linear models can be divided into the claim probability and the rate determination of the claim probability intensity (Si et al. , 2022; Wüthrich and Merz, 2022).…”

Tariff Analysis in Automobile Insurance: Is It Time to Switch from Generalized Linear Models to Generalized Additive Models?

“…Some a priori rating variables, like age, cubic capacity of the car, etc., can also be included in the model. Examples of this approach can be found in [21,22]. Nevertheless, we do not consider the Bonus-Malus Systems here.…”