Decision-Making with Complex Data Structures using Probabilistic Programming

TL;DR

This paper introduces a probabilistic programming framework for decision-making involving complex data structures like social networks and protein sequences, enabling effective reasoning in large or infinite information spaces.

Contribution

It extends decision-theoretic reasoning to complex data structures using probabilistic programming, with an approximate algorithm suitable for large or infinite information spaces.

Findings

Algorithm performs well on large information spaces

Framework supports arbitrary data types with minimal user effort

Effective decision-making in complex, real-world data scenarios

Abstract

Existing decision-theoretic reasoning frameworks such as decision networks use simple data structures and processes. However, decisions are often made based on complex data structures, such as social networks and protein sequences, and rich processes involving those structures. We present a framework for representing decision problems with complex data structures using probabilistic programming, allowing probabilistic models to be created with programming language constructs such as data structures and control flow. We provide a way to use arbitrary data types with minimal effort from the user, and an approximate decision-making algorithm that is effective even when the information space is very large or infinite. Experimental results show our algorithm working on problems with very large information spaces.