The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON COMPUTERS
Transactions ID Number: 29-117
Full Name: Hyontai Sug
Position: Associate Professor
Age: ON
Sex: Male
Address: Division of Computer and Information Engineering, Dongseo University, Busan, 617-716
Country: KOREA
Tel: 82-10-5071-7143
Tel prefix:
Fax: 82-51-327-8955
E-mail address: hyontai@yahoo.com
Other E-mails: shtdaum@hanmail.net
Title of the Paper: An Effective Sampling Method for Decision Trees Considering Comprehensibility and Accuracy
Authors as they appear in the Paper: Hyontai Sug
Email addresses of all the authors: hyontai@yahoo.com
Number of paper pages: 10
Abstract: Because the target domain of data mining using decision trees usually contains a lot of data, sampling is needed. But selecting proper samples for a given decision tree algorithm is not easy, because each decision tree algorithm has its own property in generating trees and selecting appropriate samples that represent given target data set well is difficult. As the size of samples grows, the size of generated decision trees grows with some improvement in error rates. But we cannot use larger and larger samples, because it¡¯s not easy to understand large decision trees and data overfitting problem can happen. This paper suggests a progressive approach in determining a proper sample size to generate good decision trees with respect to generated tree size and accuracy. Experiments with two representative decision tree algorithms, CART and C4.5 show very promising results.
Keywords: decision trees, proper sample size determination
EXTENSION of the file: .pdf
Special (Invited) Session: An Empirical Determination of Samples for Decision Trees
Organizer of the Session: 609-694
How Did you learn about congress:
IP ADDRESS: 118.39.140.245
