My main interests involve Reinforcement Learning and subfields of it. Additionally, i concern with domains like Focused Crawling under the perspective of reinforcement learning. In the following sections i provide some information of my current research efforts.

Ensemble Pruning

Ensemble Pruning, also known as ensemble selection, selective ensemble and ensemble thinning, deals with the reduction of the ensemble size prior to combining the members of the ensemble. It is important for two reasons: a) efficiency: Having a very large number of models in an ensemble adds a lot of computational overhead, and b) predictive performance: An ensemble may consist not only of high performance models, but also of models with lower predictive performance. Pruning the low-performing models while maintaining a good diversity of the ensemble is typically considered as a proper recipe for a successful ensemble. Publications, bibliography and source code can be found here

 

Multi-Agent Reinforcement Learning

Multi-Agent Reinforcement Learning attracts much attention in the past few years at it poses very challenging problems. Reinforcement Learning is an appealing solution to the problems that arise to Multi Agent Systems. This is due to the fact that Reinforcement Learning is a robust and well suited technique for learning in Multi Agent Systems. Recently we proposed a multi-agent Reinforcement Learning approach that uses coordinated actions which we call strategies and a voting process that combines the decisions of the agents, in order to follow a strategy.
Publications
I. Partalas, I. Feneris, I. Vlahavas. A Hybrid Multi-Agent Reinforcement Learning Approach using Strategies and Fusion. International Journal on Artificial Intelligence Tools, WorldScientific, pp. 945-962, 17 (5), 2008.
I. Partalas, I. Feneris, I. Vlahavas. Multi-Agent Reinforcement Learning using Strategies and Voting, 19th IEEE International Conference on Tools with Artificial Intelligence, pp. 318-324, Patras Greece, 2007. (pdf)

Source code can be found here. The softeware is distributed under the GNU General Public Licence. It requires Java v1.5 or better and the Pursuit Domain Package. For more information please contact Ioannis Feneris.

 

Transfer Learning

Transfer learning refers to the process of conveying experience from a simple task to another more complex (and related) task in order to reduce the amount of time that is required to learn the latter task. Typically, in a transfer learning procedure the agent learns a behavior in a source task, and it uses the gained knowledge in order to speed up the learning process in a target task. Reinforcement Learning algorithms are time expensive when they learn from scratch, especially in complex domains, and transfer learning comprises a suitable solution to speed up the training process. Publications and source code and can be found here.

Focused Crawling

Focused Crawling aims to index the web according to a specific theme and thus support domain-specific search engines and thematic web portals. Reinforcement is a very suitable approach to training focused crawlers, due to the nature of crawlers, which can only receive partial feedback at the end of a successful crawl.
Publications, bibliography, source code and datasets can be found here.