Indecision Trees: Learning Argument-Based Reasoning under Quantified Uncertainty

TL;DR

Indecision Trees are an extension of decision trees that handle uncertainty, produce probabilistic outputs, and generate logical arguments for reasoning systems, addressing real-world machine learning challenges.

Contribution

This paper introduces Indecision Trees, a novel decision tree variant that learns and infers under uncertainty while providing interpretable argument-based reasoning.

Findings

Supports inference under uncertainty

Provides a distribution over labels

Can be integrated into reasoning systems

Abstract

Using Machine Learning systems in the real world can often be problematic, with inexplicable black-box models, the assumed certainty of imperfect measurements, or providing a single classification instead of a probability distribution. This paper introduces Indecision Trees, a modification to Decision Trees which learn under uncertainty, can perform inference under uncertainty, provide a robust distribution over the possible labels, and can be disassembled into a set of logical arguments for use in other reasoning systems.