Michael Mayer

Dr. Michael Mayer

la Mobilière, Bern, Switzerland

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
Verified
Journal Article
Open access
PDF file
MMMichael Mayer
Michael Mayer
SBSteven C Bourassa
Steven C Bourassa
MHMartin Hoesli
Martin Hoesli
DSDonato Scognamiglio
Donato Scognamiglio
In some applications of supervised machine learning, it is desirable to trade model complexity with greater interpretability for some covariates while letting other covariates remain a “black box”. An important example is hedonic property valuation modeling, where machine learning techniques typically improve predictive accuracy, but are too opaque for some practical applications that require greater interpretability. This problem can be resolved by certain structured additive regression (STAR) models, which are a rich class of regression models that include the generalized linear model (GLM) and the generalized additive model (GAM). Typically, STAR models are fitted by penalized least-squares approaches. We explain how one can benefit from the excellent predictive capabilities of two advanced machine learning techniques: deep learning and gradient boosting. Furthermore, we show how STAR models can be used for supervised dimension reduction and explain under what circumstances their covariate effects can be described in a transparent way. We apply the methodology to residential land and structure valuation, with very encouraging results regarding both interpretability and predictive performance.

26 July 2023 in arXiv
Verified
Preprint
Open access
PDF file
TFTobias Fissler
Tobias Fissler
CLChristian Lorentzen
Christian Lorentzen
MMMichael Mayer
Michael Mayer
One of the main tasks of actuaries and data scientists is to build good predictive models for certain phenomena such as the claim size or the number of claims in insurance. These models ideally exploit given feature information to enhance the accuracy of prediction. This user guide revisits and clarifies statistical techniques to assess the calibration or adequacy of a model on the one hand, and to compare and rank different models on the other hand. In doing so, it emphasises the importance of specifying the prediction target functional at hand a priori (e.g. the mean or a quantile) and of choosing the scoring function in model comparison in line with this target functional. Guidance for the practical choice of the scoring function is provided. Striving to bridge the gap between science and daily practice in application, it focuses mainly on the pedagogical presentation of existing results and of best practice. The results are accompanied and illustrated by two real data case studies on workers' compensation and customer churn.

18 August 2025 in arXiv
Verified
Preprint
Open access
PDF file
MMMichael Mayer
Michael Mayer
MWMario Wüthrich
Mario Wüthrich
Originally introduced in cooperative game theory, Shapley values have become a very popular tool to explain machine learning predictions. Based on Shapley's fairness axioms, every input (feature component) gets a credit how it contributes to an output (prediction). These credits are then used to explain the prediction. The only limitation in computing the Shapley values (credits) for many different predictions is of computational nature. There are two popular sampling approximations, sampling KernelSHAP and sampling PermutationSHAP. Our first novel contributions are asymptotic normality results for these sampling approximations. Next, we show that the paired-sampling approaches provide exact results in case of interactions being of maximal order two. Furthermore, the paired-sampling PermutationSHAP possesses the additive recovery property, whereas its kernel counterpart does not.