Accompanying code for the paper "A Kernel Test for Causal Association via Noise Contrastive Backdoor Adjustment".

Last update: Nov 25, 2021

Related tags

Deep Learning kgformula

Overview

#backdoor-HSIC (bd_HSIC)

Accompanying code for the paper "A Kernel Test for Causal Association via Noise Contrastive Backdoor Adjustment".

To generate the synthetic data used in the paper, the general procedure is the following:

Generate the necessary data
Generate and run the appropriate job using the "interpreter_run.py"
Post-process the results
Generate plots from the results

We list which scripts to run and which job to run to recreate the experiments in the paper.

Experiment	File to generate data	File to run experiments	Post process file	Generate plot
Binary	generate_simple_null.py	do_null_binary	kchsic_categorical_power.py	kchsic_categorical_power.py
Linear continuous	generate_data_multivariate.py	kc_rule_new	post_process.py	create_plots.py
Mixed	generate_mixed_data.py	do_null_mix	post_process.py	create_plots.py
Non-linear continuous	generate_hdm_breaker.py	hdm_breaker_fam_y=i_100, i=1,4	post_process.py	create_plots.py
Linear kernel binary	generate_simple_null.py	do_null_binary_linear_kernel	kchsic_categorical_power.py	kchsic_categorical_power.py
Linear kernel continuous	generate_data_multivariate.py	linear	post_process.py	create_plots.py

To get the results for the benchmark post-double-selection method and the R scripts can be run. To run the R scripts, some .pt files need to be converted to .csv. To do this, use the pt_to_csv.py script. For the real world experiments, do the following:

Lalonde - Run lalonde_real_experiment.py
Twins - Process the data using twins_preprocess.py, then run twins_real_experiment.py

This repository will include the results needed to recreate the plots in the paper already.

Accompanying code for the paper "A Kernel Test for Causal Association via Noise Contrastive Backdoor Adjustment".

Related tags

Overview

Owner

Robert Hu

Pytorch implementation for Semantic Segmentation/Scene Parsing on MIT ADE20K dataset

Alfred-Restore-Iterm-Arrangement - An Alfred workflow to restore iTerm2 window Arrangements

Implementation of Diverse Semantic Image Synthesis via Probability Distribution Modeling

LightSeq is a high performance training and inference library for sequence processing and generation implemented in CUDA

[CVPR 2020] Local Class-Specific and Global Image-Level Generative Adversarial Networks for Semantic-Guided Scene Generation

CoCosNet v2: Full-Resolution Correspondence Learning for Image Translation

PyTorch implementation for Partially View-aligned Representation Learning with Noise-robust Contrastive Loss (CVPR 2021)

PRIME: A Few Primitives Can Boost Robustness to Common Corruptions

A Differentiable Recipe for Learning Visual Non-Prehensile Planar Manipulation

Unsupervised Image-to-Image Translation

[NeurIPS-2021] Slow Learning and Fast Inference: Efficient Graph Similarity Computation via Knowledge Distillation

Flask101 - FullStack Web Development with Python & JS - From TAQWA

Haze Removal can remove slight to extreme cases of haze affecting an image

GraPE is a Rust/Python library for high-performance Graph Processing and Embedding.

Face Recognition Attendance Project

Magic tool for managing internet connection in local network by @zalexdev

Kaggle Ultrasound Nerve Segmentation competition [Keras]

DrWhy is the collection of tools for eXplainable AI (XAI). It's based on shared principles and simple grammar for exploration, explanation and visualisation of predictive models.

Multiple paper open-source codes of the Microsoft Research Asia DKI group

Unofficial PyTorch Implementation for HifiFace (https://arxiv.org/abs/2106.09965)