Cognitive Behavior and Human Computer Interaction Based on Machine Learning Algorithms

8Feature Optimized Machine Learning Framework for Unbalanced Bioassays

Dinesh Kumar^1*, Anuj Kumar Sharma², Rohit Bajaj³ and Lokesh Pawar³

¹Department of CSE, Guru Kashi University, Bathinda, India

²Department of CSE, BRCMCET, MD University, Rohtak, India

³Department of CSE, Chandigarh University, Chandigarh, India

*Corresponding author: [e-mail]

Abstract

In this research, we have discussed biopsy; the biopsy is a medical procedure done by a surgeon in which a patient is tested for expansion of the disease. We applied specific machine learning espouse on a standard database to get more accurate data. The aim of the paper is found that the precise of a database is increased. Authors collect biopsy databases followed by the surgeon and perform a class balancing process, using tool name that is class balance. Then, we apply four machine learning algorithms that are Multilayer Perceptron, Bagging, Random Committee, and J48. Class Balancer is a straightforward filter to apply that weightage to every instance so that every class instance has the same value in the dataset, such that the sum of all instances in the dataset will remain unchanged. It has been observed from a specific experiment that Multilayer Perception and Bagging algorithm done excellent work, increase accuracy, decrease the error, and increase the TP (true positive rate). This research paper will provide an efficient way to increase the accuracy of the data.

Keywords: Bagging, biopsy, J48, machine learning, ...