Ciosek, KA;
(2015)
Linear Reinforcement Learning with Options.
Doctoral thesis , UCL (University College London).
Preview |
Text
phd-thesis.pdf Download (681kB) | Preview |
![[thumbnail of OptionsAggregation.zip]](https://discovery.ucl.ac.uk/style/images/fileicons/archive.png) |
Archive
OptionsAggregation.zip Download (240kB) |
Abstract
The thesis deals with linear approaches to the Markov Decision Process (MDP). In particular, we describe Policy Evaluation (PE) methods and Value Iteration (VI) methods that work with representations of MDPs that are compressed using a linear operator. We then use these methods in the context of the options framework, which is way of employing temporal abstraction to speed up MDP solving. The main novel contributions are: the analysis of convergence of the linear compression framework, a condition for when a linear compression framework is optimal, an in-depth analysis of the LSTD algorithm, the formulation of value iteration with options in the linear framework and the combination of linear state aggregation and options.
| Type: | Thesis (Doctoral) |
|---|---|
| Title: | Linear Reinforcement Learning with Options |
| Event: | UCL (University College London) |
| Open access status: | An open access version is available from UCL Discovery |
| Language: | English |
| UCL classification: | UCL > Provost and Vice Provost Offices UCL > Provost and Vice Provost Offices > UCL BEAMS UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science |
| URI: | https://discovery.ucl.ac.uk/id/eprint/1470360 |
Archive Staff Only
 |
View Item |
