FitBit and other personal trackers have become increasingly popular in recent years as people are becoming more interested in self-monitoring their personal health. Fitness Trackers are a popular area of study amongst data scientists, statisticians, medical experts, physiologists, and psychologists, just to name a few academic research areas. Detecting relationships in complex time series data, such as FitBit Fitness Tracker data, can be a way of establishing daily life patterns, and also a way of detecting deviations from these patterns. Worldwide, many insurance corporates have deliberately expressed interest in acquiring fitness trackers data. It is expected that health data from electronic sources could soon be compiled into a health or wellness report and shared with insurance companies to help them determine who they will cover.
A thorough analysis of opensource Fitbit data is performed. The key findings are highlighted and discussed. The analysis provided herein is performed using 940 data points collected from 33 distinct users. Machine Learning Models are used to solve a regression problem using Multiple Linear Regression, Random Forest and Extreme Gradient Booster
This project is associated with two medium stories including and discussing the two parts of the project, Exploratory Data Analysis and Machine Learning, respectively. If interested, you can check these two storeys at the following links: