Rule Induction by EDA with Instance-Subpopulations
نویسنده
چکیده
In this paper, a new rule induction method by using EDA with instance-subpopulations is proposed. The proposed method introduces a notion of instance-subpopulation, where a set of individuals matching a training instance. Then, EDA procedure is separately carried out for each instance-subpopulation. Individuals generated by each EDA procedure are merged to constitute the population at the next generation. We examined the proposed method on Breast-cancer in Wisconsin and Chess End-Game. The comparisons with other algorithms show the effectiveness of the proposed method.
منابع مشابه
Using Induced Rules as Complex Features in Memory-Based Language Learning
An extension to memory-based learning is described in which automatically induced rules are used as binary features. These features have an "active" value when the left-hand side of the underlying rule applies to the instance. The RIPPER rule induction algorithm is adopted for the selection of the underlying rules. The similarity of a memory instance to a new instance is measured by taking the ...
متن کاملNoise-Tolerant Rule induction from Multi-Instance data
This paper addresses the issue of multipleinstance induction of rules in the presence of noise. It first proposes a multiple-instance extensions of rule-based learning algorithms. Then, it shows what kind of noise can appear in multiple-instance data, and how to handle it theoretically. Finally, it describes the implementation of such a noise-tolerant multiple instance learner, and shows its pe...
متن کاملAdvances in Inductive Rule Learning
A extension to the k-nearest neighbor classifier is described in which automatically induced rules are used as binary features, which are active in an instance when the left-hand side of the corresponding rule matches with the instance. The ripper rule induction algorithm is employed to produce the rules. The similarity between a memory instance and a new instance is based on the rules the two ...
متن کاملPreceding Rule Induction with Instance Reduction Methods
A new prepruning technique for rule induction is presented which applies instance reduction before rule induction. An empirical evaluation records the predictive accuracy and size of rule-sets generated from 24 datasets from the UCI Machine Learning Repository. Three instance reduction algorithms (Edited Nearest Neighbour, AllKnn and DROP5) are compared. Each one is used to reduce the size of t...
متن کاملMultiple regulators control expression of the Entner-Doudoroff aldolase (Eda) of Escherichia coli.
The Escherichia coli eda gene, which encodes the Entner-Doudoroff aldolase, is central to the catabolism of several sugar acids. Here, we show that Eda synthesis is induced by growth on gluconate, glucuronate, or methyl-beta-D-glucuronide; phosphate limitation; and carbon starvation. Transcription of eda initiates from three promoters, designated P1, P2, and P4, each of which is responsible for...
متن کامل