Skip to content

anshul2209/Human-Activity-Tracker

Repository files navigation

Udacity-capstone-Project: Human Activity Tracker System

Ever wondered how smartphones or watches get to know that you have been inactive for a while or you eat a lot. The project is aimed to demystify how smartphones get to know this and build a system which can accurately detect the activity using some of the sensor data from the accelerometer and gyrometer in particular.

It is a multiclass classification problem aimed at bucketing the user activity into three buckets: (Non-hand-oriented activities, Hand-oriented activities - General, Hand-oriented activities - Eating)

Following python libraries were made us of.

  • scipy.io, os, numpy, sklearn, pandas, matplotlib, seaborn, plotly, scipy, time

Amazon sagemaker is used for model building ,training, deployment and tuning.