Decision Tree for data with Known Hierarchical Class Labels

  • Sachin Gavankar
Keywords: data mining; classification; decision tree; hierarchical class label

Abstract

Decision Tree classifier builds a classification model using training data. It consists of records having attribute values and corresponding class label. Very few algorithms deal with class labels that are organized as hierarchical structure. In this paper we propose a framework for decision trees which considering the prior knowledge of the class hierarchies in training data. Classification algorithm is applied multiple time to predict class label from higher level (coarse-grain) to lower level (fine-grain) by using respective training records.

References

[1] Jiawei Han, Michline Kamber, „Data Mining Concepts and Technique‟ , Kaufmann Pulications, 2001 [2] David Hand, Heiki Mannila, Padhraic Smyth, „Principles of Data Mining‟, The MIT Press. [3] Tom Mitchell, „Machine Learning‟, Mc-GrawHill Publications [4] J.R.Quinlan, „C4.5 Programs for Machine Learning‟, Morgan Kaufmann Publications, San Mateo, CA, 1993. [5] Yen-Liang Chen, Hsiao-Wei Hu, Kewi Tang, „Consturcting a decision tree from data with hierarchical class labels‟, Expert Systems with Applications 36 (2009) 4838-4847. [6] Alshdaifat E., Coenen F., Dures K. (2013) Hierarchical Single Label Classification: An Alternative Approach. In: Bramer M., Petridis M. (eds) Research and Development in Intelligent Systems XXX. pp-39-52 Springer, Cham. [7] I. H. Witten, E. Frank. Nuts and bolts: Machine Learning alogrithms in java. In:Data Mining: Practical Machine Learning Tools and Techniques with Java Implementation, pp.265-320. Morgan-Kaufmann, 2000. [8] Sachin Gavankar, Sudhir Sawarkar, “Decision Tree: Review of Techniques for Missing Values at Training, Testing and Compatibility”, Proc. of the 3rd Int‟l Conf on Artificial Intelligence, Modelling and Simulation, AIMS2015, Malaysia, 2015.
Published
2018-04-15
How to Cite
Gavankar, S. (2018, April 15). Decision Tree for data with Known Hierarchical Class Labels. ASIAN JOURNAL FOR CONVERGENCE IN TECHNOLOGY (AJCT ) -UGC LISTED, 4(I). https://doi.org/https://doi.org/10.33130/asian%20journals.v4iI.448
Section
Computer Science and Engineering