Data

Tools

Indexes

Applications

Experiments

Open Source Science

information-v3

Package for information-theoretic data analysis

https://github.com/vanessb/information-v3

Science Score: 44.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ✓
    CITATION.cff file
    Found CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ○
    Academic publication links
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (4.9%) to scientific vocabulary

Keywords

deep-learning entropy information-bottleneck information-theory machine-learning mutual-information
Last synced: 6 months ago · JSON representation ·

Repository

Package for information-theoretic data analysis

Basic Info
  • Host: GitHub
  • Owner: VanessB
  • License: gpl-3.0
  • Language: Jupyter Notebook
  • Default Branch: master
  • Homepage:
  • Size: 4.39 MB
Statistics
  • Stars: 5
  • Watchers: 1
  • Forks: 2
  • Open Issues: 0
  • Releases: 0
Topics
deep-learning entropy information-bottleneck information-theory machine-learning mutual-information
Created over 4 years ago · Last pushed over 2 years ago
Metadata Files
Readme License Citation

README.md

Mutinfo

Russian/Русский

An information-theoretic framework to study datasets and neural networks.

Features

  • Kernel Density Estimation (maximum likelihood and LSE) mutual information estimators.
  • Kozachenko-Leonenko (original and weighted) mutual information estimators.
  • Framework for mutual information estimation via lossy compression.
  • Synthetic datasets with predefined information-theoretic quantities.
  • Information bottleneck experiments with neural networks.

Structure

  • /source/python/mutinfo — source code of the framework, including submodules for synthetic dataset generation.
  • /source/examples — .ipynb files to demonstrate the framework and conduct experiments.
  • /source/gnuplot — gnuplot scripts to plot data acquired from experiments.

Owner

  • Login: VanessB
  • Kind: user

Citation (CITATION.cff) 

# This CITATION.cff file was generated with cffinit.
# Visit https://bit.ly/cffinit to generate yours today!

cff-version: 1.2.0
title: Information-theoretic data analysis framework
message: >-
  If you use this software, please cite it using the
  metadata from this file.
type: software
authors:
  - given-names: Ivan
    family-names: Butakov
    email: butakov.id@phsyetch.edu
    orcid: 'https://orcid.org/0000-0002-0424-6695'
  - given-names: Alexander
    family-names: Tolmachev
    email: tolmachev.ad@phystech.edu
    orcid: 'https://orcid.org/0009-0006-7711-3005'
  - given-names: Sofia
    family-names: Malanchuk
    email: malanchuk.sv@phystech.edu
  - given-names: Anna
    family-names: Neopryatnaya
    email: neopriatnaia.am@phystech.edu
  - given-names: Alexey
    family-names: Frolov
    email: al.frolov@skoltech.ru
    orcid: 'https://orcid.org/0000-0002-6734-0179'
  - email: k.andreev@skoltech.ru
    given-names: Kirill
    family-names: Andreev
    orcid: 'https://orcid.org/0000-0002-2920-2015'
repository-code: 'https://github.com/VanessB/Information-v3'
abstract: >-
  Information-theoretic data analysis framework based on the
  entropy estimation and the lossy compression. 
keywords:
  - machine learning
  - information theory
  - mutual information
  - entropy
  - information bottleneck
  - deep learning
license: GPL-3.0

GitHub Events

Total
  • Watch event: 2
Last Year
  • Watch event: 2

Dependencies

setup.py pypi
  • numpy *
  • scikit-learn *
  • scipy *