The conference “Algorithmic Learning Theory 2006” is dedicated to studies of learning from a mathematical and algorithmic perspective. Researchers consider various abstract models of the problem of learning and investigate how the learning goal in such a setting can be formulated and achieved. These models describe ways to define
– the goal of learning,
– how the learner retrieves information about its environment,
– how to form of the learner’s models of the world (in some cases).
Retrieving information is in some models is passive where the learner just views a stream of data. In other models, the learner is more active, asking questions or learning from its actions. Besides explicit formulation of hypotheses in an abstract language with respect to some indexing system, there are also more implicit methods like making predictions according to the current hypothesis on some arguments which then are evaluated with respect to their correctness and wrong predictions (coming from wrong hypotheses) incur some loss on the learner. In the following, a more detailed introduction is given to the five invited talks and then to the regular contributions.
KeywordsSupport Vector Machine Reinforcement Learning Online Learning Markov Decision Process Weak Learner
Unable to display preview. Download preview PDF.