VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.

401-3934-00L 6 Credits BSC , DR , MSC D-MATH

Data Science for Actuaries

Lecturers & Examiners: Prof. Dr. Mario Valentin Wüthrich
VVZ CR n/a

Last Updated: 2026-06-03 00:39:05

Abstract

This lecture offers a comprehensive introduction to statistical modeling and machine learning in actuarial science, covering key areas including supervised learning, unsupervised learning, and reinforcement learning.

Objective

The student is familiar with the core concepts of statistical modeling and machine learning. They understand the underlying theory, can implement these methods, and are able to compute and interpret relevant actuarial examples.

Content

We cover various topics, and all these topics are illustrated with actuarial examples. The following topics are covered: - Strictly consistent scoring - Model fitting, model validation and model selection - Regression models and model regularization - Generalized linear models - Local regression, isotonic regression - Statistical biases, balance property, auto-calibration - Gini score, Murphy's score decomposition, lift plots - Deep learning and feed-forward neural networks - Regression trees and random forests - Gradient boosting machines - Tensor data and unstructured data - Word embedding and negative sampling - Convolutional neural networks - Recurrent neural networks - Transformer architectures - Unsupervised learning (auto-encoder, clustering methods and visualizations) - Generative modeling - Variational auto-encoder - Generative-adversarial networks - Large language models (foundation models, in-context learning) - Reinforcement learning

Resources

Lecture Notes

AI Tools for Actuaries:https://aitools4actuaries.com/

General Information

Language
English
Levels
BSC , DR , MSC
Frequency
Every two years

Examination

Type
session examination
Mode
oral 30 minutes
only in person exams (i.e. no remote exams)

Course Components

Type Title Time & Place Hours
lecture with exercise Data Science for Actuaries
  • Mon 16:15-18:00 (HG E 1.2)
  • Tue 16:15-18:00 (HG D 5.2)
4 h weekly

Offered In