Using L1 penalized logistic regression to classify the activities of humans based on time series obtained by a Wireless Sensor Network.The classification tasks consist in predicting the activity performed by the user from time-series generated by a Wireless Sensor Network.
Data: This dataset contains temporal data from a Wireless Sensor Network worn by an actor performing the activities: bending, cycling, lying down, sitting, standing, walking. AReM data: link This dataset represents a real-life benchmark in the area of Activity Recognition applications.
Treatment:
- Engineered several features by using mean, first and third quartile values from sensor during the activity
- Used L1 and L2 loss to deal with sparsity
- Performed Cross Validation to find the optimum hyper parameters.
- Binary Classification (Logistic-L1) : 97% Accuracy
- Multi-class Classification (Gaussian Naive Bayes) : 81% Accuracy
Languages/Tools Used: Python, scikit-learn