A comparative study of support vector machine and logistic regression article · January 021 citations reads 11 authors
Yüklə 395,89 Kb. Pdf görüntüsü
|
|
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/368288658 A COMPARATIVE STUDY OF SUPPORT VECTOR MACHINE AND LOGISTIC REGRESSION Article · January 2021 CITATIONS 0 READS 11 2 authors , including: Jude Chukwura Obi Chukwuemeka Odumegwu Ojukwu University 15 PUBLICATIONS 6 CITATIONS SEE PROFILE All content following this page was uploaded by Jude Chukwura Obi on 03 May 2023. The user has requested enhancement of the downloaded file. O RIENTAL JOURNAL OF SCIENCE & ENGINEERING VOL -2, ISS-1, FEB - 2021 www.ojse.org ojse©2019 Page 85 A COMPARATIVE STUDY OF SUPPORT VECTOR MACHINE AND LOGISTIC REGRESSION Jude Chukwura Obi 1 Obimuanya Ijeoma 2 1 2 Department of Statistics, Chukwuemeka Odumegwu Ojukwu University, Anambra State, Nigeria Email: oj.obi@coou.edu.ng 1 Obimuanyaijeoma9@gmail.com 2 Abstract Two machine learning tools namely the Support Vector Machine (SVM) and the Logistics Regression (LR) have been studied. Both tools are binary classifiers and have been discovered to have different approaches to classification. The objective of the study includes: to find out which method is consistently better than the other; to find out why any of the two classifiers consistently performs better than the other, if this is true; to find out if specific datasets would require the use of a particular classifier than the other; to compare their accuracy; to draw conclusions based on the results of the comparison. The study adopted the secondary source of data collection which was largely sourced from internet. The statistical tools employed for the data analysis include accuracy rate, one-way analysis of variance that was used to the test the significance of the accuracy measures. The findings of the study show that there is no significant difference between the SVM Gaussian and Linear Kernel but there was significant difference between SVM Gaussian and LR. Based on the accuracy rates of the classifiers on the different datasets used in the study, the findings reveal that SVM using the Gaussian kernel was far better a classifier than LR. Yüklə 395,89 Kb. Dostları ilə paylaş:
|