Code and data sets for "Topological Machine Learning for Multivariate Time Series" (https://arxiv.org/abs/1911.12082)
Occupancy_Test1: Contains data set and codes for Test 1 (predicting future room occupancy)
Occupancy_Test2: Contains data set and codes for Test 2 (predicting past room occupancy)
Order of running code:
- dataCleaning.py
Python code for data preparation and cleaning.
- pointCloud.py
Python code for converting multivariate time series to point cloud data.
- predictionColumn.py
Python code to output a column containing room occupancy status (0 or 1) of each time window.
- TDA_distmatrix.R
R code to calculate Wasserstein distances between persistence diagrams.
- knn.py
Python code to run the k-nearest neighbors algorithm for binary classification.