Date of Publication :15th October 2019

Abstract: To manage characterization for huge information, information or data filtering and cleansing are preferred as preprocessing steps. For the most part it evacuate noisy, errors and conflicted data and results misclassification. In this paper, we performed examination of misclassified data and recognize how much information is should be redressed to get important data. To exhibit this idea, we have utilized AirTrafficDataset from Statistical Computing Statistical Graphics to analyze misclassified content in informational index. Two directed classifiers are used: Support vector Machine and decision tree. The results shows that out of these classifiers, SVM classify 85% of the data correctly and only 15% of data has misclassification.

Reference :