Action selection methods using reinforcement learning

By M. Humphreys. 1997

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

1. Diederik Marijn Roijers
2. Chunming Liu
3. Shimon Whiteson
4. J. Urgen Schmidhuber
5. Xin Xu
6. Dewen Hu
7. Peter Vamplew
8. Richard Dazeley

Citations in the corpus

• Credit Assignment in Multiple Goal Embodied Visuomotor Behavior (2010, background): “However, our focus is the embodied cognition setting, where single agent pursues multiple goals that are divided up between multiple independent modules that a single agent can activate concurrently (Humphrys, 1996; Karlsson, 1997; Singh and Cohn, 1998; Sprague and Ballard, 2003).”
• A Survey of Multi-Objective Sequential Decision-Making (2013, background): “Dusparic and Cahill (2009) extend the W-Learning algorithm of Humphrys (1996).”
• Multiobjective Reinforcement Learning: A Comprehensive Overview (2015, methodology): “One simple method to compute W values is called Top-Q [11], [32], which assigns the W value as the highest”
• HQ-Learning (1997, unknown): “Like HQ-learning, Humphrys' W-learning (1996) uses multiple Q-learning agents.”