This Project is based on the human activity detection. In this project we have collected the data of 12 activities from the sensors of the smartphone using Andro-Sensor mobile application then we have perform EDA and performed required statistical calculation and added it in our training dataset in feature engineering. We have visualise the high demensional data using t-SNE and implemented machine learning classifier SVM using poly and rbf kernel, KNN, Logistic Regression, Decision Tree , XGBoost then applied GridSearch to get best parameters and we got best accuracy in Logistic Regression and we have successfully tested new test data in model. This model can be used for health monitoring activities pattern.