Coder Social home page Coder Social logo

jurijsnazarovs / panel_me_ode Goto Github PK

View Code? Open in Web Editor NEW
8.0 1.0 1.0 8.78 MB

Official repository of the paper "A Variational Approximation for Analyzing the Dynamics of Panel Data". Mixed Effect Neural ODE. UAI 2021.

License: MIT License

Python 100.00%

panel_me_ode's Introduction

Official repository of the paper (UAI 2021) "A Variational Approximation for Analyzing the Dynamics of Panel Data", Mixed Effect Neural ODE.

Video presentation: youtube

Panel data involving longitudinal measurements of the same set of participants taken over multiple time points is common in studies to understand childhood development and disease modeling. We propose a probabilistic model called ME-NODE to incorporate (fixed + random) mixed effects for analyzing the dynamics of panel data. In the paper, we show that our model can be derived using smooth approximations of SDEs provided by the Wong-Zakai theorem. This allows us to incorporate uncertainty of the trajectory in the modelsimilar to SDE, meanwhile using ODE solvers to fit the model. In addition, we show an interesting connection to the Random projection, which provides theoretical justification of our model's ability to describe dynamics of panel data.

If you like our work, please give us a star. If you use our code in your research projects, please cite our paper as

@inproceedings{nazarovs2021variational,
  title={A variational approximation for analyzing the dynamics of panel data},
  author={Nazarovs, Jurijs and Chakraborty, Rudrasis and Tasneeyapant, Songwong and Ravi, Sathya and Singh, Vikas},
  booktitle={Uncertainty in Artificial Intelligence},
  pages={107--117},
  year={2021},
  organization={PMLR}
}

To run the main.py file, you need to download data generator repository, and either

  • keep it in a separate directory up-level from the ME-NODE repository or
  • put it in the same directory as ME-NODE repository.

Note: the access to the data genration repository is handled in lib/parse_datasets.py with sys.path.append.

Visualization of experiments

Rotating MNIST

Results of generating rotating MNIST in next time steps for different steps of calibration

TADPOLE

Results of generating brain scans in next time steps

panel_me_ode's People

Contributors

jurijsnazarovs avatar

Stargazers

Yunyi Li avatar Cuong Ngo avatar Sourav Pal avatar Joong Ho Kim avatar  avatar  avatar Ivan Zorin avatar !welcome avatar

Watchers

 avatar

Forkers

wn1695173791

panel_me_ode's Issues

Big issue: too cool

This is very cool work. Thank you for your contribution to the scientific community!

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.