Watkins and Dayan: convergence of Q-learning

What changed

Tabular model-free action-value learning acquired a stated convergence guarantee under specified conditions.

Scientific scope

The 1992 paper's theorem and assumptions; not invention solely in 1992 or convergence of nonlinear deep Q networks.

Evidence

The linked sources identify this contribution and its scope. WRS has not independently reconstructed the scientific claims. Watkins outlined the algorithm in his 1989 thesis; this record is specifically the 1992 two-author convergence paper and does not claim general deep-RL convergence.

Contributors

• Christopher J. C. H. Watkins · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
• Peter Dayan · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.

• Watkins and Dayan, Q-learning (1992)

Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.

Full scientific record · Browse contribution grades · Request a correction