statistical-learning-theory

[ICML2026] The first, fully verified, sorry-free, large-scale Lean 4 library for statistical learning theory, covering infrastructures for mordern statistics and learning theory.

Testing VC dimension & Rademacher complexity generalization error bounds for a simple perceptron and a rectangular classifier.

This repository contains the paper and artifact for: "Finite-Budget Structural Identifiability under Bounded Observation" The official archived version of the paper is available on Zenodo: https://doi.org/10.5281/zenodo.18736348

Machine learning course notes covering statistical ML, generalization theory and optimization, and modern temporal learning—from classical models to transformers.

Verification code for block-term operator theory: why rank-(L,L,1) block-term neural operators generalise better than CP, Tucker and TT at matched capacity. Four scripts, no data or GPU needed.