Mahidol University Logo
Faculty of ICT, Mahidol University
 

Admissions

Printable Version

 

ACCURACY COMPARISON AMONG CLASSIFICATION TECHNIQUES OF OPEN SOURCE DATA MINING SOFTWARE

 

TITLE ACCURACY COMPARISON AMONG CLASSIFICATION TECHNIQUES OF OPEN SOURCE DATA MINING SOFTWARE
AUTHOR RAVADEE BUATHONG
DEGREE MASTER OF SCIENCE PROGRAMME IN COMPUTER SCIENCE
FACULTY FACULTY OF SCIENCE
ADVISOR SONGSRI TANGSRIPAIROJ
CO-ADVISOR SUDSANGUAN NGAMSURIYAROJ
DAMRAS WONGSAWANG
 

ABSTRACT

The purpose of this research project was to study the abilities of two types of open source data mining software, which are Weka and Rapid Miner. In particular, accuracy comparison among classification techniques of these two software programs was investigated. Five classification algorithms, including Decision Tree, Naïve Bayes, Neural Network, Rule Induction, and K-Nearest Neighbor were focused on. Eight data sets from the UCI Machine Learning Repository website were used for classification model evaluation. These data sets were selected based on a different number of attributes, records, and classes. Two evaluation methods of classification model were used: percentage split and 10 - fold cross-validation. Only Weka was tested with the percentage split method by dividing the data set into the training set, in proportions of 50% ,60%, 70% ,80% ,and 90% ,and the test set in proportions of 50%, 40%, 30%, 20% ,and 10%. Both Weka and Rapid Miner were tested with the 10-fold cross- validation method. The accuracy of the classification model of each classification techniques was measured as evaluation results and presented in the form of a comparative table and graphs.


KEYWORD OPEN SOURCE DATA MINING SOFTWARE / CLASSIFICATION TECHNIQUE / ACCURACY

 

 

Go to Top

 

ICT Building, Mahidol University, 999 Phuttamonthon 4 Road, Salaya, Nakhonpathom 73170 Tel. +66 02 441-0909 Fax. +66 02 849-6099
Mahidol University Computing Center, The Faculty of ICT, Mahidol University , Rama 6 Road, Rajathevi, Bangkok 10400 Tel. +66 02 354-4333 Fax. +66 02 354-7333