Giter Club home page Giter Club logo

batchmultivalidconformal's Introduction

Batch Multivalid Conformal Prediction

This repository contains code for the paper Batch Multivalid Conformal Prediction by Christopher Jung, Georgy Noarov, Ramya Ramalingam, and Aaron Roth.

The src/ folder contains Python implementations for our multivalid conformal predictors, along with other utilities such as plotting tools and conformal scorers. In particular:

  1. The algorithm BatchMVP is implemented in /src/MultivalidAlgorithms/MultivalidCoverage.py
  2. The algorithm BatchGCP is implemented in /src/MultivalidAlgorithms/GroupCoverage.py

The experiments/ folder contains Jupyter notebooks implementing the experiments in the paper.

batchmultivalidconformal's People

Contributors

progbelarus avatar ramya-ramalingam avatar

Stargazers

Wei-Chen.Chen avatar avatar Afroditi Papadaki avatar Katina Mooneyham avatar Madhav Jha avatar avatar avatar avatar valeman avatar Yaodong Yu avatar avatar

Watchers

Varun Gupta avatar avatar avatar

Forkers

valeman

batchmultivalidconformal's Issues

Variance mismatch between paper and notebook in synthetic experiment

Hello,
I was looking at the code in your SyntheticExperiment notebook, specifically at lines

std_dev = np.dot(xs_binvars, std_dev_list) + y_std
ys = np.dot(xs, theta) + np.random.normal(loc=0, scale= std_dev, size=T)

The variance of the Gaussian error produced by this code is $(\sum_i x_i\sigma_i + \sigma)^2$, which is different from the variance $\sum_i x_i\sigma_i^2 + \sigma^2$ as written in section 5.1 of the paper because of the cross-terms. I'd suggest to fix the version on ArXiv, or recompute the results and see if this makes any difference.

Recommend Projects

  • React photo React

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

  • Vue.js photo Vue.js

    🖖 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. 📊📈🎉

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 ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.