Weightless Swarm Algorithm as Transformation Technique for Improving the Performance of Machine Learning Algorithms
N. Bharanidharan, Sai Siva Sasank T, Jeevan Sreeram Reddy, Naga Sujan T · 2021 5th International Conference on Trends in Electronics and Informatics (ICOEI) · 2021
Nowadays machine learning algorithms are applied to various fields but still there is a requirement for improving the accuracy of machine learning algorithms for certain applications. This paper focuses on using Weightless Swarm Algorithm as transformation technique for enhancing the accurateness of machine learning techniques in SONAR dataset classification. Three different machine learning techniques namely Random Forest, Stochastic Gradient Descent, and Decision Tree are examined as classifier for categorizing the SONAR data as either mineral or normal rock. Notably, Mathews correlation Coefficient of Random forest classifier is 75.27% and this is increased to 81.81% through the usage of Weightless Swarm Algorithm as transformation technique.