Travelled to:1 × Australia
1 × Canada
1 × USA
Collaborated with:R.Schoknecht
Talks about:converg (3) reinforc (2) function (2) approxim (2) learn (2) necessari (1) algorithm (1) synchron (1) gradient (1) provabl (1)
Person: Artur Merke
DBLP: Merke:Artur
Contributed to:
200420032002Wrote 3 papers:
- ICML-2004-MerkeS #approximate #convergence #learning #linear
- Convergence of synchronous reinforcement learning with linear function approximation (AM, RS).
- ICML-2003-SchoknechtM #algorithm #performance
- TD(0) Converges Provably Faster than the Residual Gradient Algorithm (RS, AM), pp. 680–687.
- ICML-2002-MerkeS #approximate #convergence #learning
- A Necessary Condition of Convergence for Reinforcement Learning with Function Approximation (AM, RS), pp. 411–418.