arrow_back Big Data Structures
Emerging patterns of lifestyle impact on personal health & wellness
Submitted by Tanmay Gupta (@tanmaygupta) on Sunday, 10 April 2016
Lifestyle is changing at a very rapid pace as we enter the internet era. Pace of evolution in terms of technology, lifestyle, work environment, etc. is more rapid than ever before and has resulted in how our lifestyle and health has changed. To be able to understand the new health and wellness patterns emerging, and help a preventive health care based start-up design improved solutions to help people preserve their health, I conducted a study using data collected by their team to understand some of the relationships between lifestyle, common medical complaints faced by people and their inter-relationships. The correlation analysis performed on the dataset identified various connections between different health related problems. These results were then visualized to draw insights using network analysis, and communities amongst the different connections were identified using modularity index. The whole study presented some very interesting relationships like regular stress and lack of sleep being closely associated with a combination of heartburn, gastritis & headaches etc. and many more similar findings.
This talk will present some of the interesting facts derived from handshake between healthcare & machine learning techniques. Few questions that would be answered are:
Does our lifestyle has any effect on our health?
We often hear people talking about health and fitness and how the modern lifestyle has affected the way we live. But how many of us are truly serious about our health? Health is largely impacted by a combination of lifestyle, environmental factors, genetic make-up an individual and the kind of care the individual receives.
How the real time data was collected and refined?
We conducted a survey and asked over 4,000 people to fill in a questionnaire which included an individual’s demographic, lifestyle, medical complaints, family history, and even gynecology details. The population comprised of 70% males & 30% females with mean population age of 35 years.
What does correlation analysis tells about health conditions?
One of the clusters in the network analysis depicted a close connection between comorbidities viz. heartburn, gastritis, Hhead aches, sleep & stress which a clear showed an association between long working hours, unhealthy food habits and low physical activity of today’s working force.
What should be our next steps?
The study was able to identify a series of correlation between various lifestyle patterns and clusters in the young population, to be able to develop more effective preventive /mitigation strategies for some of the new age health problems.
This session would present insights on health related problems using machine learning techniques for everyone. So, the only requirement would be a basic understanding of network analysis (as in social network), a curiosity to learn and open mind for a healthy diccussion.
Tanmay is a researcher with a background in application of machine learning algorithms to product development & enhancement. His specialties include predictive analytics & mathematical modeling using R as the primary tool. He has developed predictive models on patient readmission, length of stay, cost model etc. and published case study & research papers in International Journal of Medical Science and Public Health (IJMSPH), Harvard Business Journal (HBJ) & International Journal of Science & Research (IJSR). He has also presented some of his work at the annual American Medical Informatics Conference (AMIA) as well.
He has more than 9 years of industry experience with 6+ years in healthcare domain. He did his B.Tech in Electronics Engineering and pursued professional course in Business Analytics & Intelligence from IIM-B.