• Karthik S
• Manjushree K

Big data mining can be referred as the methodology of inferring meaningful information from large datasets. The type of algorithms used for mining information from such data influences the efficiency and performance of the overall process. Classification is one of the well-known problems under the domain of data mining. It attempts to classify the data into multiple groups. There are variety of data mining algorithms available for classification such as naïve Bayesian, Fuzzy logic, Latent Dirichlet Allocation, Decision trees etc. One such model which possess reasonably better accuracy for majority of the cases is the Support Vector Machines. The reason for the success of SVM lies in the fact that users can choose an appropriate kernel function according to the complexity of relationship model. In this work, Pearson VII Universal Kernel (PUK) function based Support vector machine is implemented for chosen datasets and its performance compared with various other classification models. Experimental results show that for the completely two different datasets that are used in this work, PUK based SVM outperforms all other classification algorithms with an accuracy of 100% followed by RBF based SVM achieving 100% and 99.51% respectively for breast cancer dataset and bank information dataset.

Data mining, SVM, Pearson VII kernel, Classification

"STUDY OF PERFORMANCE OF SVM ALGORITHM BASED ON PEARSON UNIVERSAL KERNEL FOR DATA CLASSIFICATION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 6, page no.k293-k298, June-2022, Available :http://www.jetir.org/papers/JETIR2206A40.pdf

"STUDY OF PERFORMANCE OF SVM ALGORITHM BASED ON PEARSON UNIVERSAL KERNEL FOR DATA CLASSIFICATION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 6, page no. ppk293-k298, June-2022, Available at : http://www.jetir.org/papers/JETIR2206A40.pdf