Learn More
In this paper, we present algorithms that perform gradient ascent of the average reward in a partially observable Markov decision process (ÈÇÅÅÈ). These algorithms are based on ÈÇÅÅÈ, an algorithm introduced in a companion paper (Baxter & Bartlett, 2001), which computes biased estimates of the performance gradient in ÈÇÅÅÈs. The algorithm's chief advantages(More)
In this paper we present TDLEAF(), a variation on the TD() algorithm that enables it to be used in conjunction with game-tree search. We present some experiments in which our chess program " KnightCap " used TDLEAF() to learn its evaluation function while playing on Internet chess servers. The main success we report is that KnightCap improved from a 1650(More)
During evolution, the amino acid sequence of a protein is much more variable and changes more rapidly than its tertiary structure. Given sufficient time, the amino acid sequences of proteins derived from a common precursor may alter to the point that they are no longer demonstrably homologous. The ability to make meaningful comparisons between such(More)
In this paper we present TDLeaf(λ), a variation on the TD(λ) algorithm that enables it to be used in conjunction with game-tree search. We present some experiments in which our chess program " KnightCap " used TDLeaf(λ) to learn its evaluation function while playing on the Free Internet Chess Server (FICS, fics.onenet.net). The main success we report is(More)
The three-dimensional structure of goose-type lysozyme (GEWL), determined by x-ray crystallography and refined at high resolution, has similarities to the structures of hen (chicken) egg-white lysozyme (HEWL) and bacteriophage T4 lysozyme (T4L). The nature of the structural correspondence suggests that all three classes of lysozyme diverged from a common(More)
The modes of binding to thermolysin of two phosphonamidate peptide inhibitors, carbobenzoxy-GlyP-L-Leu-L-Leu (ZGPLL) and carbobenzoxy-L-PheP-L-Leu-L-Ala (ZFPLA), have been determined by X-ray crystallography and refined at high resolution to crystallographic R-values of 17.7% and 17.0%, respectively. (GlyP is used to indicate that the trigonal carbon of the(More)
Dots-and-Boxes is a child's game which remains analytically unsolved. We implement and evolve ar-tiicial neural networks to play this game, evaluating them against simple heuristic players. Our networks do not evaluate or predict the nal outcome of the game, but rather recommend moves at each stage. Superior generalisation of play by co-evolved populations(More)