Methods to Handle Multiclass Imbalance Data in Educational Data Mining
Bhasha Anjaria1, Ankita Gandhi2, Jay Gandhi3
1Bhasha Anjaria*, lecturer, Information Technology, Parul Polytechnic Instituute, Vadodara, Gujarat.
2Ankita Gandhi, Deputy HOD & an Assistant Professor in Computer Science and Engineering Department,PIET, Parul University Vadodara, Gujarat, India.
3Jay Gandhi, Assistant Professor in Computer Science and Engineering Department,PIET, Parul University Vadodara, Gujarat, India.
Manuscript received on April 18, 2020. | Revised Manuscript received on July 22, 2020. | Manuscript published on April 30, 2020. | PP: 654-657 | Volume-9 Issue-4, April 2020. | Retrieval Number: D7571049420/2020©BEIESP | DOI: 10.35940/ijeat.D7571.049420
Open Access | Ethics and Policies | Cite | Mendeley
© 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: In Scientists ordinarily exclude the equalization of the dissemination on a dataset in Educational Data Mining (EDM). It can truly influence the consequence of the classification procedure. Hypothetically, the distribution of data is respectively balanced pretended by the majority of classifier. Hence, the execution of the classification algorithm simply turned out to be less viable and should be taken care of the issue could illuminated. These exploration would characterize about imbalanced class on multiclass EDM dataset minding component utilizing the Map Reduce. This strategy serves adjusting system for the dataset’s dissemination, using parallel processing; those classification result will the results. These balancing strategies can be implemented with different kind of classification methods like Naïve Bayes, SVM, NN to measure the improvisation in the results.
Keywords: Educational Datamining, Imbalanced class classification, MapReduce, Multiclass, Resampling Techniques.