The ICS artificial intelligence group is also distinguished by its concern with machine learning -- the mechanisms by which knowledge is acquired through experience. This requires attention to a variety of problems: forming concepts, learning search heuristics, learning how to decompose a problem into subproblems, improving motor skills, explaining new examples in terms of existing knowledge, revising and extending existing theories in light of new evidence, discovering scientific laws and theories. The ICS AI group has active research projects in all of these areas.

The group is also concerned with moving beyond the isolated machine learning systems that have traditionally been studied, toward producing integrated systems with learning components that interact in significant ways. The focus on learning incorporates issues of representation and performance, since one cannot construct an AI learning system that ignores these aspects. The ICS AI group is therefore simultaneously exploring alternative schemes for representing knowledge -- rule-based systems, frame-like structures, probabilistic concepts, and neural networks -- along with processes that employ these different notations. The group's focus on learning serves to constrain these representations and algorithms, since not all approaches support the modularity and adaptability required by learning systems that improve over time.

You can receive further information on the following topics:

• The Machine Learning Database Repository
• Some machine learning programs
• Machine learning papers from UCI
• Archive issues of the Machine Learning List
• Other information related to machine learning

[ICS Home | UCI Home]