Inductive Logic Programming: 15th International Conference, by Hiroki Arimura, Takeaki Uno (auth.), Stefan Kramer, Bernhard

By Hiroki Arimura, Takeaki Uno (auth.), Stefan Kramer, Bernhard Pfahringer (eds.)

1 “Change is inevitable.” Embracing this quote we've attempted to scrupulously exp- iment with the structure of this convention, the fifteenth overseas convention on Inductive good judgment Programming, with a bit of luck making it even larger than it already was once. however it can be as much as you, the inquisitive reader of this booklet, to pass judgement on our luck. the key adjustments comprised broadening the scope of the convention to incorporate extra diversified sorts of non-propositional studying, to once more have tutorials on fascinating new parts, and, for the ?rst time, to actually have a discovery problem as a platform for collaborative paintings. This yr the convention was once co-located with ICML 2005, the twenty second Inter- tional convention on desktop studying, and in addition in shut proximity to IJCAI 2005, the nineteenth overseas Joint convention on Arti?cial Intelligence. - position may be difficult, yet we enormously bene?ted from the neighborhood help supplied via Codrina Lauth, Michael may well, and others. We have been additionally in a position to invite all ILP and ICML contributors to shared occasions together with a poster consultation, an invited speak, and an instructional concerning the fascinating new quarter of “statistical relational lea- ing”. extra invited talks have been solely given to ILP members and have been provided as a type of stock-taking—?ttingly so for the fifteenth occasion in a series—but additionally attempted to supply a recipe for destiny endeavours.

Show description

Read or Download Inductive Logic Programming: 15th International Conference, ILP 2005, Bonn, Germany, August 10-13, 2005. Proceedings PDF

Similar international conferences and symposiums books

Object-Oriented Programming: 8th European Conference, ECOOP '94 Bologna, Italy, July 4–8, 1994 Proceedings

This quantity comprises the lawsuits of the eighth eu convention on Object-Oriented Programming (ECCOP '94), held in Bologna, Italy in July 1994. ECOOP is the most advantageous eu occasion on object-oriented programming and know-how. The 25 complete refereed papers awarded within the quantity have been chosen from 161 submissions; they're grouped in classes on classification layout, concurrency, styles, declarative programming, implementation, specification, dispatching, and adventure.

Principles of Data Mining and Knowledge Discovery: First European Symposium, PKDD '97 Trondheim, Norway, June 24–27, 1997 Proceedings

This e-book constitutes the refereed lawsuits of the 1st ecu Symposium on ideas of knowledge Mining and data Discovery, PKDD '97, held in Trondheim, Norway, in June 1997. the quantity provides a complete of 38 revised complete papers including abstracts of 1 invited speak and 4 tutorials.

Interactive Distributed Multimedia Systems and Telecommunication Services: 5th International Workshop, IDMS'98 Oslo, Norway, September 8–11, 1998 Proceedings

This e-book constitutes the refereed court cases of the fifth overseas Workshop on Interactive dispensed Multimedia structures and Telecommunication providers, IDMS'98, held in Oslo, Norway, in September 1998. The 23 revised complete papers offered have been conscientiously chosen from a complete of sixty eight submissions.

Current Trends in Database Technology – EDBT 2006: EDBT 2006 Workshops PhD, DataX, IIDB, IIHA, ICSNW, QLQP, PIM, PaRMA, and Reactivity on the Web, Munich, Germany, March 26-31, 2006, Revised Selected Papers

This publication constitutes the completely refereed joint post-proceedings of 9 workshops held as a part of the tenth foreign convention on Extending Database know-how, EDBT 2006, held in Munich, Germany in March 2006. The 70 revised complete papers awarded have been chosen from quite a few submissions in the course of rounds of reviewing and revision.

Additional resources for Inductive Logic Programming: 15th International Conference, ILP 2005, Bonn, Germany, August 10-13, 2005. Proceedings

Sample text

Because the state space S of all possible states st can be intractably large, the classical tabular representations of Q(st , a) for Q learning or V (st ) for value learning are impractical. Inspired by the MAXQ framework [7], we introduce a value decomposition that expresses the Q function in terms of V values: Q(st , a) = V (u) + V (ua ) , (2) u∈st where ua denotes the concept instance inferred by action a. While making the problem tractable, this approximation establishes a relationship between the 26 N.

16) (is-block A) (is-block B) (is-block C) 0 (0) 0 (0) 0 (0) Cycle 2 Fig. 2. An example of value propagation for three successive cycles in the blocks world. Updated values are shown in parentheses whereas others have been used to guide inference. The literals in grey have been inferred to be false. argument in (3) to the set of instances that depend directly on u. This reflects the intuition that the values of instances which depend on u are more indicative of the desirability of inferring u than the values of other instances in the fringe.

In a different approach, Cohen and Singer [15] used bootstrapping to learn similar rules. These approaches assumed a query-based, top-down inference mechanism and sought to modify the logic program itself in order to achieve performance gains. In our work, however, we consider a data-driven, bottom-up inference process that is more appropriate for reactive agents, for the reasons discussed earlier. In addition, our approach operates on top of a fixed logic program and modifies the way it is utilized by the logic interpreter.

Download PDF sample

Rated 4.86 of 5 – based on 37 votes