dpny518 / human-activity-recognition Goto Github PK
View Code? Open in Web Editor NEWThis project forked from dusanrandd/human-activity-recognition
Human activity recognition is very important tool in health monitoring of patients suffering from various chronic diseases and for professional and recreational sportsman. In this repository will be presented two AI models that classify data given from Smartwatch Accelerometer of observed persons. Classification is consists of ten chosen daily living activities from WISDM database. AI models that are used are Deep Neural Network and Random Forest Classifier. Results of both models are presented and it shows that they have similar results on authentication persons that were in the training data group and recognition of activities of persons that are used only in test group of the dataset.
License: GNU General Public License v3.0