Paper Info
Title
Recurrent neural networks with backtrack-points and negative reinforcement applied to cost-based abduction
Abstract
Abduction is the process of proceeding from data describing a set of observations or events, to a set of hypotheses which best explains or accounts for the data. Cost-based abduction (CBA) is an AI formalism in which evidence to be explained is treated as a goal to be proven, proofs have costs based on how much needs to be assumed to complete the proof, and the set of assumptions needed to complete the least-cost proof are taken as the best explanation for the given evidence. In this paper, we present two techniques for improving the performance of high order recurrent networks (HORN) applied to cost-based abduction. In the backtrack-points technique, we use heuristics to recognize early that the network trajectory is moving in the wrong direction; we then restore the network state to a previously stored point, and apply heuristic perturbations to nudge the network trajectory in a different direction. In the negative reinforcement technique, we add hyperedges to the network to reduce the attractiveness of local minima. We apply these techniques to a suite of six large CBA instances, systematically generated to be difficult.
Year
DOI
Venue
2005
10.1016/j.neunet.2005.06.026
Neural Networks
Keywords
Field
DocType
Reasoning under uncertainty,Belief revision,cost-based abduction,Bayesian networks,high-order networks,Recurrent networks
Heuristic,Mathematical optimization,Computer science,Recurrent neural network,Maxima and minima,Bayesian network,Heuristics,Mathematical proof,Wrong direction,Artificial intelligence,Trajectory,Machine learning
Journal
Volume
Issue
ISSN
18
5
0893-6080
Citations 
PageRank 
References 
6
0.48
16
Authors
4
Name
Order
Citations
PageRank
Ashraf M. Abdelbar124325.43
Mostafa A. El-Hemaly260.48
Emad A. M. Andrews3182.74
Wunsch II Donald C.4135491.73