Formalization of a Stochastic Approximation Theorem

1 code implementation • 12 Feb 2022 • Koundinya Vajjha, Barry Trager, Avraham Shinnar, Vasily Pestun

Stochastic approximation algorithms are iterative procedures which are used to approximate a target value in an environment where the target is unknown and direct observations are corrupted by noise.

CertRL: Formalizing Convergence Proofs for Value and Policy Iteration in Coq

1 code implementation • 23 Sep 2020 • Koundinya Vajjha, Avraham Shinnar, Vasily Pestun, Barry Trager, Nathan Fulton

Reinforcement learning algorithms solve sequential decision-making problems in probabilistic environments by optimizing for long-term reward.

Decision Making reinforcement-learning +1

Verification of ML Systems via Reparameterization

no code implementations • 14 Jul 2020 • Jean-Baptiste Tristan, Joseph Tassarotti, Koundinya Vajjha, Michael L. Wick, Anindya Banerjee

Proof assistants can be used to formally verify machine learning systems by constructing machine checked proofs of correctness that rule out such bugs.

BIG-bench Machine Learning Fairness +1

A Formal Proof of PAC Learnability for Decision Stumps

no code implementations • 1 Nov 2019 • Joseph Tassarotti, Koundinya Vajjha, Anindya Banerjee, Jean-Baptiste Tristan

We present a formal proof in Lean of probably approximately correct (PAC) learnability of the concept class of decision stumps.

BIG-bench Machine Learning Learning Theory