Advanced Topics in Artificial Intelligence: 12th Australian by Geoffrey Holmes, Mark Hall, Eibe Prank (auth.), Norman Foo

By Geoffrey Holmes, Mark Hall, Eibe Prank (auth.), Norman Foo (eds.)

The twelfth Australian Joint convention on synthetic Intelligence (AI'QQ) held in Sydney, Australia, 6-10 December 1999, is the newest in a sequence of annual re­ gional conferences at which advances in man made intelligence are mentioned. This sequence now draws many overseas papers, and certainly the structure of this system committee displays this geographical variety. along with the standard tutorials and workshops, this 12 months the convention integrated a spouse sympo­ sium at which papers on commercial appUcations have been awarded. The symposium papers were released in a separate quantity edited by way of Eric Tsui. Ar99 is equipped by way of the college of latest South Wales, and subsidized by means of the Aus­ tralian laptop Society, the Commonwealth medical and commercial learn business enterprise (CSIRO), machine Sciences company, the KRRU staff at Griffith college, the Australian man made Intelligence Institute, and Neuron- Works Ltd. Ar99 acquired over a hundred and twenty convention paper submissions, of which approximately o- 3rd have been from outdoors Australia. promenade those, 39 have been authorized for normal presentation, and one more 15 for poster show. those complaints comprise the total general papers and prolonged summaries of the poster papers. All papers have been refereed, normally by way of or 3 reviewers chosen by means of participants of this system committee, and an inventory of those reviewers appears to be like later. The technical software comprised days of workshops and tutorials, fol­ lowed by means of 3 days of convention and symposium plenary and paper sessions.

Show description

Read or Download Advanced Topics in Artificial Intelligence: 12th Australian Joint Conference on Artificial Intelligence, AI’99 Sydney, Australia, December 6–10, 1999 Proceedings PDF

Similar nonfiction_7 books

A chronology of the United States Marine Corps, Volume III : 1947-1964

This is often the 3rd quantity of a chronology of Marine Corps actions which covers the heritage of the U. S. Marines. it truly is derived from professional documents and acceptable released ancient works. This chronology is released for the knowledge of all attracted to Marine Corps actions in the course of the interval 1947 - 1964 and is devoted to these Marines who participated within the occasions indexed.

Unifying Petri Nets: Advances in Petri Nets

When you consider that their creation approximately forty years in the past, learn on Petri nets has diverged in lots of diverse instructions. a number of periods of Petri internet, inspired both by means of idea or functions, with its personal particular beneficial properties and techniques of study, were proposed and reviews extensive. those winning advancements have resulted in a really heterogeneous panorama of numerous versions, and this, in flip, has motivated study on suggestions and techniques that give a contribution to unifying and structuring the varied panorama.

Deep Blue: An Artificial Intelligence Milestone

S a competitor of the Deep Blue workforce, I had combined feelings as I A watched their chess-playing desktop defeat international Chess Cham­ pion Garry Kasparov in the course of their 1997 Rematch. at the one hand, it intended that our MIT software, *Socrates, wouldn't be the 1st software to defeat a human international Chess Champion.

Extra resources for Advanced Topics in Artificial Intelligence: 12th Australian Joint Conference on Artificial Intelligence, AI’99 Sydney, Australia, December 6–10, 1999 Proceedings

Sample text

It is important for our learning methods that enough learning iterations are allowed. Results obtained using only 1/5 as many such iterations showed no evidence of learning, even though we were using ten times as many linear equations per learning iteration. We used search depth 3 because it was the highest we could afford, given the computational resources available and the number of experiments. Some experiments were conducted for search depths other than 3. Depth 1 was clearly too small. Learning with depth 1 gave improved play for Chess and Draughts, but made play much worse for Lose-chess and Game 3.

This shows that there are n shots, with shot 3 having m examples in its description. 2 Age Agel Agc2 V>hu2 A^ Value n AniAutrZam Avenge Viluc A^l A|t2 Vahiel Vahicl Attribute Z o o B Domiiuiil Motinn Vilucl VltlBl Age Agel A r l Value B Value m AltttNiK Length Age Vahwl Age! AgeJ Vahwm Age. Fig. 3. A typical description hierarchy used by the incremental learning algorithm. Each shot description has an age and 9 attributes and each one of the attributes has a set of data values-l-age values associated with it.

3. D. H. Cron, Directing Search in Metagame, M. Comp. thesis. School of Computer Science and Software Engineering, Monash University, to appear. 4. S. L. Epstein, The intelligent novice — learning to play better, in: D. N. L. Levy and D. F. ). Heuristic Programming in Artificial Intelligence: The First Computer Olympiad, EUis Horwood, Chichester, 1989. 5. S. L. Epstein, J. Gelfand and J. Lesniak, Pattern-based learning and spatieilly oriented concept formation in a multi-agent, decision-making expert, Computational Intelligence 12 (1996) 198-221.

Download PDF sample

Rated 4.69 of 5 – based on 7 votes