Abstract

Predefined category exists for text categorization. In a document, text may be of any type category like government, education or health etc. many methods exist in market invented by researchers for text categorization. One of them is k-NN (k nearest neighbor) algorithm. k play a role to define number of classes for categorization. A training set is generated for each type of category to check its performance than whole text categorized. There is a problem of missing information during training sets. After study recent years invention on k-NN, we find out a solution of this problem. Multiple-Level Learning will improve the performance of k-NN. So in this paper we study about k-NN and propose hybrid algorithm with combination of Multiple-Level Learning and k-NN.

References

• E. Fix, and J. Hodges, "Discriminatory analysis. Nonparametric discrimination: Consistency properties". Technical Report 4, USAF School of Aviation Medicine, Randolph Field, Texas, 1951
• T. M. Cover, and P. E. Hart, "Nearest neighbor patternclassification", IEEE Transactions on Information Theory, 13, pp. 21–27, 1967
• R. O. Duda, and P. E. Hart, Pattern classification andscene analysis, New York: Wiley, 1973.
• W. Yu, and W. Zhengguo, "A fast kNN algorithm for textcategorization", Proceedings of the Sixth International Conference on Machine Learning and Cybernetics, Hong Kong, pp. 3436-3441, 2007
• W. Yi, B. Shi, and W. Zhang'ou, "A Fast KNN Algorithm Applied to Web Text Categorization", Journalof The China Society for Scientific and TechnicalInformation, 26(1), pp. 60-64, 2007.
• K. G. Anil, "On optimum choice of k in nearest neighborclassification", Computational Statistics and DataAnalysis, 50, pp. 3113–3123, 2006
• E. Kushilevitz, R. Ostrovsky, and Y. Rabani, "Efficientsearch for approximate nearest neighbor in highdimensional spaces". SIAM Journal on Computing, 30, pp. 457–474, 2000
• M. Lindenbaum, S. Markovitch, and D. Rusakov, "Selective sampling for nearest neighbor classifiers", Machine Learning, 54, pp. 125–152, 2004, IJCSI International Journal of Computer Science Issues, Vol. 7, Issue 4, No 2, July 2010 21www. IJCSI. org
• C. Zhou, Y. Yan, and Q. Chen, "Improving nearestneighbor classification with cam weighted distance". Pattern Recognition, 39, pp. 635–645, 2006
• D. P. Muni, and N. R. D. Pal, "Genetic programming forsimultaneous feature selection and classifier design",IEEE Transactions on Systems Man and Cybernetics PartB – Cybernetics, 36, pp. 106–117, 2006.
• J. Neumann, C. Schnorr, and G. Steidl, "Combined SVMbasedfeature selection and classification", Machine Learning, 61, pp. 129–150, 2005
• Tanya Taneja, Balraj Sharma, "Text Classification Using PSO & Other Technique" International Journal of Recent Development in Engineering and TechnologyWebsite: www. ijrdet. com (ISSN 2347-6435(Online) Volume 3, Issue 1, July 2014)
• Zhang W. , Yoshida T. , and Tang X," Text classification using multi-word features", In proceedings of the IEEE international conference on Systems, Man and Cybernetics, pp. 3519 – 3524, 2007.
• Witten, I. H. and Frank, E. "Data Mining: Practical Machine Learning Tools and Techniques", (2nd Edition), Morgan Kaufmann, San Francisco, 2005.