The applications of data mining and machine learning in Bangladesh for disease pattern analysis and prediction

Abstract

Over the years, data mining and machine learning have proved to be very convenient in numerous fields of science and technology and their applications in the medical sector is an emerging one. With the world population rate increasing by the year, the medical sector is generating immense amount of data every day. By storing this data and analyzing it for disease patterns, using numerous data mining and machine learning techniques, predictive models can be built to assess future risk to potential patients. These models may have a very important role in a developing country like Bangladesh, where Non-Communicable Diseases (NCD) like diabetes and heart diseases have affected a large portion of its population. Clinical diagnosis of these diseases requires a lot of tests which complicates the prediction process and proves to be expensive for most patients as well. Predictive models based on data mining and machine learning techniques provides a much more efficient system of predicting future risks for patients, saving lives and a lot of money. This project looks at several data mining and machine learning techniques for analyzing medical data in order to recognize disease patterns, compare their performances and eventually produces a model with the highest accuracy in disease prediction.