Get Started
Home
Topics
Search
Library
Research questionHow should model architecture and size scale with data for noisy tabular ratemaking?Actuarial ratemaking uses heterogeneous, noisy tabular data, where adding examples or parameters may produce uneven predictive gains. The practical problem is deciding which scaling choices improve out-of-sample count predictions.
AI
Finance
Machine Learning
Statistical Machine Learning
Latest papersRecent research connected to this question, newest first.Scaling Laws, Tabular Data and Actuarial Ratemaking ModelsThe evidence comes from a real-world motor-insurance portfolio, comparing multiple model families across training-data fractions and random seeds using out-of-sample Poisson deviance. It examines TabM, supervised tabular Transformers, MLP baselines, and related objective or inductive-bias choices; broader generalization beyond this portfolio is not established.research paper · Sep 2, 2026
Related questions
How should tabular foundation models be adapted for censored time-to-event prediction under competing risks?Can tabular foundation models learn transferable physical laws with units and noiseless mechanisms, not just interpolate data?How can reliable foundation-model scaling laws be constructed without training every configuration?How can general-purpose tabular predictors balance adaptation-free accuracy with computational cost?