An implementation of "Stochastic Diffusion: A Diffusion Based Model for Stochastic Time Series Forecasting" (KDD '25).
Jupyternote book stochdiff provides an instruction to run the model and make forecasting on the Cochlear patient data.
You can train the model with your own datasets with the train.py.
The codes are written in pytorch 1.13, please setup your environment correspondingly to run the model.
If you find this model useful and put it in your publication, we encourage you to add the following references:
@inproceedings{10.1145/3711896.3737137,
author = {Liu, Yuansan and Wijewickrema, Sudanthi and Hu, Dongting and Bester, Christofer and O'Leary, Stephen and Bailey, James},
title = {Stochastic Diffusion: A Diffusion Based Model for Stochastic Time Series Forecasting},
year = {2025},
isbn = {9798400714542},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3711896.3737137},
doi = {10.1145/3711896.3737137},
pages = {1939–1950},
numpages = {12},
location = {Toronto ON, Canada},
series = {KDD '25}
}