Machine Learning

Medical Big Data

With the provided electronic health records and demographic analysis of a large population, machine learning can be applied to correlate health of an individual to various parameters (variables), both extrinsic and intrinsic. The medical records consist of fields, of type, categorical, character and numerical. The field variables represent the quantitative values belonging to different groups, which determine the health of an individual such as, physiological parameters, diseases existing in family history, eating habits like consumption of healthy food, which includes meat, vegetables, etc., unhealthy habits like drinking, smoking, social status determined by the state of job, marriage, etc., healthy habits like having balanced diet, exercising regularly. The aim of this project is to conduct detail analysis of data as per given objective for optimal allocation of medical resources for longevity and to guide in changing one's life style to avoid complications in long run by developing descriptive and predictive models. Various algorithms will be evaluated using a cross-validation approach to get better prediction accuracy.

File Code 1 : ARIMA model

Predicting future values from 2014 to 2020 based on values from 2002 to 2013