#This is a readme file on the feature extraction tasks from different multiple sensor data.
#There are six physical factors of a room collected, and there are 16 different activities to be predicted.
#A list of weather events was also added to the building dataset, and there are 22 different features in the input list.
#We applied the discretization and one-hot encoding on the input, and we considered the prediction of six different physical features given the activity type.
#Afterward, we applied linear regression, LASSO, Support Vector Regression, Gradient Boosted Regression Trees.
#For the classification, we applied Support Vector Machine, LSTM, feed-forward DNN, and Hybrid DNN.
#Code is written in python 3.5 with tensorflow 1.6