MultiVerse: Causal Reasoning using Importance Sampling in Probabilistic Programming (2019-10-17T00:00:00.000000Z)

TL;DR

By considering the structure of the counterfactual query, one can significantly optimise the inference process, and MultiVerse, a probabilistic programming prototype engine for approximate causal reasoning, is introduced.

Abstract

We elaborate on using importance sampling for causal reasoning, in particular for counterfactual inference. We show how this can be implemented natively in probabilistic programming. By considering the structure of the counterfactual query, one can significantly optimise the inference process. We also consider design choices to enable further optimisations. We introduce MultiVerse, a probabilistic programming prototype engine for approximate causal reasoning. We provide experimental results and compare with Pyro, an existing probabilistic programming framework with some of causal reasoning tools.

Authors

Jonathan G. Richens

2 papers

Ciarán M. Lee

2 papers

Saurabh Johri

2 papers

TL;DR

Abstract

Authors

References62 items

Counterfactual diagnosis

Meaningful Explanations of Black Box AI Decision Systems

Counterfactual Off-Policy Evaluation with Gumbel-Max Structural Causal Models

Universal Marginalizer for Amortised Inference and Embedding of Generative Models

Simple, Distributed, and Accelerated Probabilistic Programming

Pyro: Deep Universal Probabilistic Programming

Woulda, Coulda, Shoulda: Counterfactually-Guided Policy Search

An Introduction to Probabilistic Programming

Conversational Explanations of Machine Learning Predictions Through Class-contrastive Counterfactual Statements

A design proposal for Gen: probabilistic programming with fast custom inference via code generation

Local Rule-Based Explanations of Black Box Decision Systems

Fairness in Decision-Making - The Causal Explanation Formula

Offline A/B Testing for Recommender Systems

Theoretical Impediments to Machine Learning With Seven Sparks from the Causal Revolution

When Worlds Collide: Integrating Different Counterfactual Assumptions in Fairness

A Universal Marginalizer for Amortized Inference in Generative Models

Counterfactual Explanations Without Opening the Black Box: Automated Decisions and the GDPR

Counterfactual Data-Fusion for Online Reinforcement Learners

Counterfactual Learning from Bandit Feedback under Deterministic Logging : A Case Study in Statistical Machine Translation

Counterfactual Multi-Agent Policy Gradients

Reliable Decision Support using Counterfactual Models

Counterfactual Fairness

Using synthetic data to train neural networks is model-based reasoning

Deep Amortized Inference for Probabilistic Programs

Inference Compilation and Universal Probabilistic Programming

Design and Implementation of Probabilistic Programming Language Anglican

Applications of Probabilistic Programming (Master's thesis, 2015)

Learning Representations for Counterfactual Inference

Inference Networks for Sequential Monte Carlo in Graphical Models

Data-driven Sequential Monte Carlo in Probabilistic Programming

Probabilistic Programming in Anglican

Neural Adaptive Sequential Monte Carlo

Counterfactual Estimation and Optimization of Click Metrics in Search Engines: A Case Study

MADE: Masked Autoencoder for Distribution Estimation

Counterfactual Risk Minimization

Probabilistic programming

Venture: a higher-order probabilistic programming platform with programmable inference

A New Approach to Probabilistic Programming Inference

A Compilation Target for Probabilistic Programming Languages

Using the Propensity Score Method to Estimate Causal Effects

Counterfactual reasoning and learning systems: the example of computational advertising

Causal Inference Using Graphical Models with the R Package pcalg

Lightweight Implementations of Probabilistic Programming Languages Via Transformational Compilation

Church: a language for generative models

Using the Probabilistic Logic Programming Language P-log for Causal and Counterfactual Reasoning and Non-Naive Conditioning

Recognition Networks for Approximate Inference in BN20 Networks

Survey Sampling

Probabilistic Evaluation of Counterfactual Queries

Probabilistic reasoning in intelligent systems: Networks of plausible inference

A Language for Counterfactual Generative Models

Causal modeling in machine learning

Lecture notes for causality in machine learning, section 9

Pyro documentation (version of the 12th of October 2019, accessed on the 13th of October, 2019)

2016; Le et al., 2017) and in particular importance sampling inference (Douglas et al., 2017; Walecki et al., 2018) can be used to facilitate the posterior inference problem

Value of information analysis for interventional and counterfactual Bayesian networks in forensic medical sciences

Single World Intervention Graphs ( SWIGs ) : A Unification of the Counterfactual and Graphical Approaches to Causality

Exploiting conditional independence for efficient, automatic multicore inference for Church

Under Consideration for Publication in Theory and Practice of Logic Programming Cp-logic: a Language of Causal Probabilistic Events and Its Relation to Logic Programming

Under Consideration for Publication in Theory and Practice of Logic Programming Probabilistic Reasoning with Answer Sets

Under consideration for publication in Theory and Practice of Logic Programming

Causality : Models , Reasoning , and Inference

Main page of the 6th Workshop on probabilistic logic programming (accessed on the 21st of

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names