Preprocesing pipline for generating features from raw HAR data
-
Batch: generation of features from original files in whole batches
-
Online: processing sample by sample and generating features on the fly
-
Both modes include test data
It is possible to generate more fatures than in the original UCI-Dataset
Based on: https://github.com/anas337/Human-Activity-Recognition-Using-Smartphones.github.io/