Skip to content

Dpm-a/ML-Human-Activity-Recognition

Repository files navigation

The aim of the project was to apply ML and Data Mining algorithms to recognise an activity performed by a subject through data captured by a smartphone.The datasets was analyzed through Exploratory Data Analysis applying the techniques of Di- mensionality Reduction, Anomaly Detection, and Imbalance Learning. Continuing with classification algorithms (e.g. Neural Networks, SVM, Ensemble Classifiers, Linear and Logistic Regression) and applying advanced and transactional clustering algorithms (e.g. OPTICS, X-Means, TX-Means). Furthermore the data was explored through was time series analysis and sequential pattern mining. Lastly, the results obtained were explained using Explainable AI