Urinary Tract Infection Analysis using Machine Learning based Classification and ANN- A Study of Prediction

Abstract

Urinary tract infection is the most frequently diagnosed infection among humans. A urinary tract infection (UTI) affects the areas of urinary system which includes the ureters, bladder, kidneys and urethra. The primary infected area of urinary system involves the lower tract i.e. bladder and urethra. The infection in bladder is painful as well as uncomfortable but if it spreads to kidneys, it can have severe consequences. Women are more susceptible to urinary infection in comparison to men due to their physiology. This paper aims to study and assess the impact and causes of urinary tract infection in human beings and evaluate the machine learning approach for urinary disease forecasting. The paper also proposed machine learning based methodology for the prediction of the urinary infection and estimating the outcomes of the designed procedures over real-time data and validating the same. The paper focuses to get high prediction accuracy of UTI using confusion matrix by Machine Based Classification and ANN technique. Some specific parameters have been selected with the help of Analysis of variance technique. The naive bayes classifier, J48 decision tree algorithm, and Artificial neural network have been used for the prediction of presence of urinary infection. The accuracy achieved by the proposed model is 95.5% approximately.