next up previous
Next: 6.1 Limitations in the Up: Accelerating Reinforcement Learning by Previous: 5.2 Performance Variation with

6 Limitations


Limitations come in , roughly, two kinds: those arising from the overall approach and those arising from the way it was implemented. In the former case, ways to address these limitations may be highly speculative, or impossible without abandoning some of the fundamental ideas behind the approach. In the latter case, there is a reasonable expectation that future work will address these limitations. The following sections will deal with these cases in turn.

Chris Drummond
Thursday January 31 01:30:31 EST 2002