SMCL - Stochastic Model Checker for Learning in Games

TL;DR

This paper introduces SMCL, a stochastic model checker designed to analyze and compare the short-term performance of various game-theoretic learning algorithms, aiding in selecting suitable algorithms for practical distributed decision-making.

Contribution

The paper presents a novel stochastic model checking method that automates comparison of learning algorithms' behavior, utilizing a new behaviour-similarity relation to handle large state spaces efficiently.

Findings

Effective comparison of learning algorithms demonstrated on multiple examples.

Automated and tunable for accuracy and speed.

Assists in selecting appropriate algorithms for distributed decision-making.

Abstract

A stochastic model checker is presented for analysing the performance of game-theoretic learning algorithms. The method enables the comparison of short-term behaviour of learning algorithms intended for practical use. The procedure of comparison is automated and it can be tuned for accuracy and speed. Users can choose from among various learning algorithms to select a suitable one for a given practical problem. The powerful performance of the method is enabled by a novel behaviour-similarity-relation, which compacts large state spaces into small ones. The stochastic model checking tool is tested on a set of examples classified into four categories to demonstrate the effectiveness of selecting suitable algorithms for distributed decision making.