ERIC Number: ED353955
Record Type: Non-Journal
Publication Date: 1992-Jan
Reference Count: N/A
COMPOSER: A Probabilistic Solution to the Utility Problem in Speed-up Learning.
Gratch, Jonathan; DeJong, Gerald
In machine learning there is considerable interest in techniques which improve planning ability. Initial investigations have identified a wide variety of techniques to address this issue. Progress has been hampered by the utility problem, a basic tradeoff between the benefit of learned knowledge and the cost to locate and apply relevant knowledge. In this paper we describe the COMPOSER system. COMPOSER embodies a probabilistic solution to the utility problem. It is implemented in the PRODIGY architecture. We compare COMPOSER to four other approaches which appear in the literature: (1) PRODIGY/EBL's Utility Analysis; (2) STATIC's Nonrecursive Hypothesis; (3) DYNAMIC: A Composite System; and (4) PALO's Chernoff Bounds. (Contains 24 references.) (Author/ALF)
Publication Type: Information Analyses; Reports - Research
Education Level: N/A
Sponsor: National Science Foundation, Washington, DC.
Authoring Institution: Illinois Univ., Urbana. Dept. of Computer Science.