Empirical Evaluation of Various Classification Methods
Shelja Sharma, Veena Mittal, Ritesh Srivastava, Sanjeev Kumar Singh · 2020 2nd International Conference on Advances in Computing, Communication Control and Networking (ICACCCN) · 2020
Humans are anxious to know how computer will venture on data, citation of teaching a computer about the data in order to get relieved from the programming at every step by the programmer and computer itself is able to perform those tasks that could not be done manually. It gives us the glimpse of machine learning. Machine learning uses various methods, statistical techniques and algorithms have the skill to understand multidimensional and multi variety of data. In this paper, we discuss about the empirical evaluation of various classification methods on the performance measure such as accuracy, kappa statistics, mean absolute errors working on different benchmarks, datasets. The classification methods we have taken in this paper are Naïve Bayesian, logistic regression, k-Star; sequential minimal optimization and random tree and the datasets are breast cancer, zoo and iris. This helps us too know which classifier is suited best for the concern objective and the parameters on which we have analyzed are accuracy, kappa statistics and absolute mean error.