Search

Recursive partitioning, also known as decision trees and classification and regression trees (CART), is a machine learning procedure that has gained traction in the behavioral sciences because of its ability to search for nonlinear and interactive effects, and produce interpretable predictive models. The recursive partitioning algorithm is greedy—searching for the variable and the splitting value that maximizes outcome homogeneity. Thus, the algorithm can be overly sensitive to chance associations in the data, particularly in small samples. In an effort to limit chance associations, we propose and evaluate a reliability-based cost function for recursive partitioning. The reliability-based cost function increases the likelihood of selecting variables that are more reliable, which should have more consistent associations with the outcome of interest. Two reliability-based cost functions are proposed, evaluated through simulation, and compared to the CART algorithm. Results indicate that reliability-based cost functions can be beneficial, particularly with smaller samples and when more reliable variables are important to the prediction, but can overlook important associations between the outcome and lower reliability predictors. The use of these cost functions was illustrated using data on depression and suicidal ideation from the National Longitudinal Survey of Youth. [ABSTRACT FROM AUTHOR]

Thanks to nonparametric estimators coming from machine learning, microlevel reserving has become more and more popular for actuaries. Recent research focused on how to integrate the whole information one can have on claims to predict individual reserves, with varying success due to incomplete observations. Using the CART algorithm, we develop new results that allow us to deal with large reporting delays and partially observed explanatory variables. Statistically speaking, we extend CART to take into account truncation of the data and introduce plug-in estimators. Our applications are based on real-life insurance portfolios embedding Income Protection and Third-Party Liability guarantees. The full knowledge of the claim lifetime is shown to be crucial to predict the individual reserves efficiently. [ABSTRACT FROM AUTHOR]