Optimized Predictive Models in Health Care Using Machine Learning

7An Optimized Feature-Based Prediction Model for Grouping the Liver Patients

Bhupender Yadav¹* and Rohit Bajaj²

¹Om Sterling Global University, Hisar, Haryana, India

²Chandigarh University, Mohali, Punjab, India

Abstract

Liver illnesses have given an insight into electronic health records and reports that include patient medical information and various other disorders. These analytics must, however, be reflected and integrated if models involving physiological pathways are to be introduced. Machine learning algorithms are applied to predict the number of healthy liver patients. This paper focuses on predicting the illness early on so we can take the necessary actions to cure it. Prediction on consistent data will lead to better performance. The paper is focused on applying machine learning techniques and a data mining tool to uncover previously unfamiliar features of the data. However, it was noticed that some variables are irrelevant in predicting more accurate outcomes, and there is an opportunity for improvement. Gini Index is used to overcome this discrepancy, and our dataset is exposed to superior feature selection. It calculates the probability of a specific feature being classed incorrectly when randomly selected. Machine learning algorithms were again applied, which resulted in greater accuracy in predicting the number of fit liver patients. Following the successful execution, the best algorithm out of all the executed algorithms is chosen as an output. Random forest ...