With the advancement of new technologies in our emerging world, Every part of our life has become an Activity. Monitoring Human activities has been an active research area since past few years with the growth of increasing demands in Health Sector. This research helps to detect Emergency situations and provides quick aid for aged people. Many Sensor based Approaches have been introduced such as Accelerometer, Gyroscope. This research paper covers predicting Activity of human using Decision Trees and Random Forest. It also discusses advantages and disadvantages of mentioned Sensor technologies. We have considered Various activities such as Running, walking, Laying, Standing etc. in our present proposed model. Firstly, Paper starts with discussing problem of the occupational diseases and Preventing People from these diseases. Then the collected data is trained and evaluated by ML Techniques. The trained model shows a satisfactory performance in all the stages. Finally, a recognition system has been developed with an accuracy of 93% in Random Forest Classifier. Experimental results showed that compared to Decision Tree Classifier, Random Forest Classifier predicts better over these various activities.

Keywords : Machine Learning , Decision Trees , Random Forest , Classification , Sensors.