Thumbnail
Access Restriction
Open

Author Hu, Yuh-Jyh ♦ Kibler, Dennis
Source CiteSeerX
Content type Text
File Format PDF
Language English
Subject Domain (in DDC) Computer science, information & general works ♦ Data processing & computer science
Subject Keyword Constructive Induction Word Count ♦ Relative Gain Measure ♦ Subsymbolic Learner ♦ Representational Bias ♦ Representational Inadequacy ♦ Standard Machine Learning Algorithm ♦ New Constructive Induction Algorithm ♦ Real-valued Attribute ♦ Wrapper Approach ♦ Small Number ♦ Inductive Algorithm ♦ Real Domain ♦ Learning Algorithm ♦ Gala Preprocessor ♦ Constructive Induction ♦ New Boolean Attribute ♦ Symbolic Learner ♦ Present Result ♦ Previous Research
Description In AAAI-96
Inductive algorithms rely strongly on their representational biases. Representational inadequacy can be mitigated by constructive induction. This paper introduces the notion of a relative gain measure and describes a new constructive induction algorithm (GALA) which is independent of the learning algorithm. GALA generates a small number of new boolean attributes from existing boolean, nominal or real-valued attributes. Unlike most previous research on constructive induction, our methods are designed as preprocessing step before standard machine learning algorithms are applied. We present results which demonstrate the effectiveness of GALA on both artificial and real domains for both symbolic and subsymbolic learners. For symbolic learners, we used C4.5 and CN2. For subsymbolic learners, we used perceptron and backpropagation. In all cases, the GALA preprocessor increased the performance of the learning algorithm. 1 A Wrapper Approach for Constructive Induction Word Count...
Educational Role Student ♦ Teacher
Age Range above 22 year
Educational Use Research
Education Level UG and PG ♦ Career/Technical Study
Learning Resource Type Article
Publisher Date 1996-01-01