Model interpretability and understanding for PyTorch

Features selector based on the self selected-algorithm, loss function and validation method

XAI - An eXplainability toolbox for machine learning

Leave One Feature Out Importance

Using / reproducing ACD (ICLR 2019) from the paper "Hierarchical interpretations for neural network predictions"

This package can be used for dominance analysis or Shapley Value Regression for finding relative importance of predictors on given dataset. This library can be used for key driver analysis or marginal resource allocation models.

Code for using CDEP from the paper "Interpretations are useful: penalizing explanations to align neural networks with prior knowledge" https://arxiv.org/abs/1909.13584

Routines and data structures for using isarn-sketches idiomatically in Apache Spark

Using / reproducing DAC from the paper "Disentangled Attribution Curves for Interpreting Random Forests and Boosted Trees"

A Julia package for interpretable machine learning with stochastic Shapley values

CancelOut is a layer for deep neural networks, that can help identify a subset of relevant input features for streaming or static data.

A Rust implementation of fANOVA (functional analysis of variance)

An R package for computing asymmetric Shapley values to assess causality in any trained machine learning model

Analyzing the Features which leads to heart diseases and visualizing the models' performance and important features using eli5, shap and pdp.

Implementation of various feature selection methods using TensorFlow library.

Variance-based Feature Importance in Neural Networks

Predicting the ideological direction of Supreme Court decisions: ensemble vs. unified case-based model

OncoNetExplainer: Explainable Prediction of Cancer Types Based on Gene Expression Data

Feature Selection

Extract textual information from Amazon products reviews and draw correlations through regression and fluctuation analysis.

Feature importance by the permutation method (for fastai V1)

Using / reproducing TRIM from the paper "Transformation Importance with Applications to Cosmology"

We find which attributes matter the most for loan decisions using neural attention. NIPS 2019 workshop

OCaml wrapper to the R randomForest package

The given information of network connection, model predicts if connection has some intrusion or not. Binary classification for good and bad type of the connection further converting to multi-class classification and most prominent is feature importance analysis.

Used the Global Terrorism Database to Explore Features of Suicide Bombings

Analysis of 'Attention is not Explanation' performed for the University of Amsterdam's Fairness, Accountability, Confidentiality and Transparency in AI Course Assignment, January 2020

This is a custom library for data processing, visualization and machine learning tools.

Given enough data, could we make predictions on whether a terrorist attack will be successful, or not? This analysis aims to do just that using Decision Trees and Random Forests created with scikit-learn. (Python)