ALACPD exploits an LSTM-autoencoder-based neural network to perform unsupervised online CPD; it continuously adapts to the incoming samples without keeping the previously received input, thus being memory-free.
To run the code with the desired specifications you can use "run.sbatch" file.
To evaluate our model, we have used datasets offered by Turing Change Point Dataset.
Red lines depict the detected change-points:
This code has been tested on Python 3.6 using the following libraries:
tensorflow 1.14.0
numpy 1.19.2
scikit-learn 0.24.2
The start of this implementation are LSTNet and OED.
email: z.atashgahi@utwente.nl