This volume contains the papers that were presented
at the Third Workshop on Algorithmic Learning Theory (ALT '92),
which was held at the CSK Information Education Center in Tokyo
from October 20 to 22, 1992.
In addition to 3 invited papers,
this volumne contains 19 papers accepted for presentation at the workshop.
The contributions in the proceedings were selected from 29 extended abstracts submitted in response to the call for papers, at the final selection meeting of the program committee held in Tokyo on June 19, 1992. The volumne contains three invited papers: ``Discovery Learning in Intelligent Tutoring Systems'' (by S. Otsuki), ``From Inductive Inference to Algorithmic Learning Theory'' (by R. Wiehagen), and ``A Stochastic Approach to Genetic Information Processing'' (by A. Konagaya).
By now the importance of machine learning to the success of the next generation of AI systems has been widely recognized and accepted. At the same time, decades of theoretical research in inductive inference and its complexity-theoretic analogue have led to the emergence of algorithmic respectively computational learning theory. ALT is the Japanese series of international workshops focusing on these learning-theoretical issues. The ALT workshops have been held annually since 1990, and are organized and sponsored by the Japanese Society for Artificial Intelligence (JSAI). The main objective of these workshops is to provide an forum for discussions and exchanges of ideas between researchers from various backgrounds in this emerging, interdisciplinary field of learning theory.
There are two concurrent series, AII initiated in 1986 in Europe, and COLT started in 1988 in the USA. It is our intention to integrate the international community of scientists interested in algorithmic respectively computational learning theory. A first step towards such an integration is to exchange all available information. We are grateful to Springer-Verlag for providing an opportunity to present the proceedings of ALT'92 to a wider international community. A further step may be to integrate even learning theory conferences for a higher concentration of scientific discussions and more efficient transfer of ideas between disciplines.
The editors are deeply grateful to all the program committee members and referees who took part in the evaluation and selection of submitted papers. In particular, we wish to thank M. Numao, T. Shinohara, and Y. Takada for their excellent work. The program committee thanks all three invited lectures for having accepted the invitation.
We thank all those who made this workshop possible, especially K. Miura, T. Nishino, and A. Sakurai. Finally, we also wish to express our gratitude to CSK for the assistance and support with local arrangements.
Tokyo, September, 1993
S. Doshita (Kyoto Univ.)
LOCAL ARRANGEMENTS COMMITTEE
Japanese Society for Artificial Intelligence (JSAI)
Information Processing Society of Japan (IPSJ)