Mobile Solution for data mining and decision support: Weight monitoring and early prediction of cardiac arrest.

Abstract

The daily accumulation of data through various means has led to the popularity of data mining in recent times. Through the use of the data mining techniques, data that are collected are used for problem-solving and other purposes. In data mining, patterns and trends of large datasets are studied using computer-based techniques. This thesis is using an Android mobile application as a data sampling tool for data mining purposes. Using this application, a predictive machine learning model, was developed to predict the probability of occurrence of cardiac of arrest in users of a mobile app over a ten- year span. The designed mobile application also functions as a support tool for weight management and fitness. The mobile application was connected to a real-time database and a machine learning tool using a Python program to perform prediction. The machine learning was based on Logistic Regression that is one of the predominant models used in the healthcare sector for predictions. The system used the user’s age, height, weight, activity level and diabetes status to predict the user’s chances of getting a Sudden Cardiac Arrest (SCA) over a ten-year period. A detailed account of the implementation processes and principles are discussed throughout this work.