Information about the opponent is essential to improve automated negotiation strategies for bilateral multiissue negotiation. In this paper we propose a negotiation strategy that e...
Koen V. Hindriks, Catholijn M. Jonker, Dmytro Tykh...
Rules are commonly used for classification because they are modular, intelligible and easy to learn. Existing work in classification rule learning assumes the goal is to produce ca...
Gradient Boosted Regression Trees (GBRT) are the current state-of-the-art learning paradigm for machine learned websearch ranking — a domain notorious for very large data sets. ...
Stephen Tyree, Kilian Q. Weinberger, Kunal Agrawal...
This paper describes a genetic learning system called SIA, which learns attributes based rules from a set of preclassified examples. Examples may be described with a variable numbe...