Indecision Trees: Learning Argument-Based Reasoning under Quantified Uncertainty
This work addresses the need for interpretable and uncertainty-aware machine learning models, particularly for applications requiring reasoning under uncertainty, though it appears incremental as it modifies existing decision trees.
The paper tackles the problem of using machine learning in real-world settings by addressing issues like black-box models, assumed certainty of imperfect measurements, and lack of probability distributions, introducing Indecision Trees that learn and perform inference under uncertainty, provide robust label distributions, and can be disassembled into logical arguments.
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.