Explaining Deep Learning Decision using LIME for Sensor-based Activity Recognition

A novel approach is discussed here to recognize human activities from body sensor information using
Reccurent Neural Network (RNN) and explaining the model decisions using Local Interpretable
Model-Agnostic Explanations (LIME). The raw signals are trained using a deep learning model consists
of RNN. The trained RNN is then used to recognize different human activities and then LIME is applied
on the model to explain the model decision for testing data.