Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Performance Assessment of Different Classification Techniques


Affiliations
1 Department of Computer Science and Applications, SCSVMV University, Enathur, India
     

   Subscribe/Renew Journal


In Data mining, Classification of data is very emblematic task. Given a vector of attributes, to predict the value of selected discrete class variable correctly is the goal of Classification. The main objective of this work is from the performance evaluation of different classifiers of data mining techniques to find the best classifier. A study has been conducted during December 2016-January 2017 with different category respondents of 1023. The questionnaire was designed to collect the factors about common cyber crime threats among the various sectors respondents in and around Chennai. For the purpose of concluding best classifier among different classifiers, the WEKA tool is used. In this study, J48, BayesNet, RandomForest, Logistic classifiers were used for the purpose of analyzing the best classifier for the cyber crime dataset. The Classification technique that has latent to extensively advance the common or predictable methods will be recommended for use in many sectors.

Keywords

BayesNet, Classification, J48, Kappa Statistic, Logistic, RandomForest, Simulation error, WEKA.
User
Subscription Login to verify subscription
Notifications
Font Size

Abstract Views: 152

PDF Views: 3




  • Performance Assessment of Different Classification Techniques

Abstract Views: 152  |  PDF Views: 3

Authors

K. Chitra Lekha
Department of Computer Science and Applications, SCSVMV University, Enathur, India
S. Prakasam
Department of Computer Science and Applications, SCSVMV University, Enathur, India

Abstract


In Data mining, Classification of data is very emblematic task. Given a vector of attributes, to predict the value of selected discrete class variable correctly is the goal of Classification. The main objective of this work is from the performance evaluation of different classifiers of data mining techniques to find the best classifier. A study has been conducted during December 2016-January 2017 with different category respondents of 1023. The questionnaire was designed to collect the factors about common cyber crime threats among the various sectors respondents in and around Chennai. For the purpose of concluding best classifier among different classifiers, the WEKA tool is used. In this study, J48, BayesNet, RandomForest, Logistic classifiers were used for the purpose of analyzing the best classifier for the cyber crime dataset. The Classification technique that has latent to extensively advance the common or predictable methods will be recommended for use in many sectors.

Keywords


BayesNet, Classification, J48, Kappa Statistic, Logistic, RandomForest, Simulation error, WEKA.