This is a source code for
T. Omi, N. Ueda, and K. Aihara, "Fully neural network based model for general temporal point processes", Advances in Neural Information Processing Systems 32 (Neurips 2019), 2120 (2019).
Paper:
[Arxiv] https://arxiv.org/abs/1905.09690
[GitHub] https://omitakahiro.github.io/files/slide/Omi2019_Neurips.pdf
NOTE: Scripts in this repository don't work for the latest version of TensorFlow. Please see the branch python_setup for how to setup the environment, which was contributed by @mattclifford1
Please see the keras_implementation.ipynb for how to implement the proposed model with Keras. More details can be found in code.ipynb
The jupyter notebook can be also viewed at https://nbviewer.jupyter.org/github/omitakahiro/NeuralNetworkPointProcess/blob/master/keras_implementation.ipynb https://nbviewer.jupyter.org/github/omitakahiro/NeuralNetworkPointProcess/blob/master/code.ipynb