A Reduced Error Pruning Technique for Improving Accuracy of Decision Tree Learning
Rinkal Patel1, Rajanikanth Aluvalu2
1Rinkal Patel, Computer Engineering, R.K University, Rajkot, India.
2Rajanikanth Aluvalu, Computer Engineering, R.K University, Rajkot, India.
Manuscript received on May 22, 2014. | Revised Manuscript received on June 04, 2014. | Manuscript published on June 30, 2014. | PP: 8-11 | Volume-3, Issue-5, June 2014. | Retrieval Number: E3056063514/2013©BEIESP
Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Decision tree inductions are well thought-out as it is one of the most accepted approaches for representing classifiers. Many researchers from varied disciplines like Statistics, Pattern Reorganization; Machine Learning measured the problem of growing a decision tree from available data. Databases are the rich sources of hidden information that can be used for intelligent decision making. Classification and Prediction techniques of data mining are two form of data analysis that can be used to discovering this type of hidden knowledge. Classification techniques deal with categorical attributes whereas prediction model is the continuous value function. Training data are analyzed by classification algorithm. In decision tree construction attribute selection measure are used to select attributes, that best partition tuples into different classes. The branches of decision tree may reflect noise or outliers in training data. So tree pruning techniques applies to identify and remove those branches which reflect noise with the aim of improving classification accuracy. But still scalability is the issue of decision tree from large database. This paper present implementation of decision tree induction algorithm in java with reduced error pruning(REP) technique for improving accuracy of classifier.
Keywords: Data Mining, Classification Decision Tree Induction, Information Gain, C4.5, Tree Pruning.