Sparse Mixture of Hidden Markov Models for Graph Connected Entities
Source code for the implementation of SpaMHMM, as described in the paper:
D. Pernes and J. S. Cardoso, "SpaMHMM: Sparse Mixture of Hidden Markov Models for Graph Connected Entities," 2019 International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, 2019, pp. 1-10. doi: 10.1109/IJCNN.2019.8851929 URL: http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8851929&isnumber=8851681.
Full text (accepted version) available here.
The experiments described there may be reproduced by running the scripts wifi_test.py
and h36m_test.py
after downloading the
respective datasets (see informations below).
Wi-Fi dataset: the dataset may be downloaded from this link. If you use this dataset, please cite the following reference:
Anisa Allahdadi, Ricardo Morla, and Jaime S. Cardoso. "802.11 wireless simulation and anomaly detection using HMM and UBM". CoRR, abs/1707.02933, 2017. URL http://arxiv.org/abs/1707.02933.
Human3.6M dataset: preprocessed data can be downloaded from this link (third party provider). Please do not forget to check the dataset license agreement, available at the Human3.6M dataset website.