From numerical proportions to analogical proportions between probabilities

2026-06-22 · Source: Takara TLDR - Daily AI Papers · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Mathematics & Computational Sciences · Depth: Expert, medium

Summary

Analogical proportions link four items a, b, c, d by a relation stating that "a is to b as c is to d", a, b, c, d being the formal representation of real world entities, ranging from simple numerical values to more complex structures such as profiles. Henri Prade and Gilles Richard's paper extends this concept to a probabilistic setting, studying analogical proportions between probability values or distributions, while ensuring normalization is preserved. The authors investigate properties of definitions based on arithmetic proportion, or a combination of arithmetic and geometric proportions. Building on prior work showing that componentwise analogical proportions in vector-represented profiles can predict class analogies, this study similarly associates each profile with a distribution describing discrete attribute frequencies. It then experimentally examines if distributions associated with four analogically proportional profiles also form an analogical proportion.

Key takeaway

For research scientists exploring novel reasoning paradigms or probabilistic modeling, this work suggests a new avenue for applying analogical reasoning. You should consider integrating the proposed probabilistic analogical proportions, based on arithmetic or combined arithmetic/geometric definitions, into your classification or distribution comparison tasks. This could enhance models by capturing relational structures within frequency distributions, potentially leading to more accurate predictions in complex data environments.

Key insights

The paper extends analogical proportions from traditional data types to probabilities and distributions.

Principles

• Analogical proportions link four items (a, b, c, d).
• Componentwise analogical proportions can predict class analogies.
• Probabilistic settings require normalization preservation.

Method

The study investigates analogical proportions between probabilities or distributions, examining definitions based on arithmetic or combined arithmetic/geometric proportions, and experimentally testing if associated distributions maintain proportionality.

In practice

• Apply analogical proportions to probabilistic data.
• Use for classification based on distribution analogies.

Topics

• Analogical Reasoning
• Probability Distributions
• Arithmetic Proportion
• Geometric Proportion
• Classification Methods
• Probabilistic Modeling

Best for: AI Scientist, Research Scientist

Related on AIssential

• LOTTERY: Learning from Reference-Only Samples in Two-Sample Testing under Size Asymmetry — LOTTERY enables robust two-sample testing in few-shot, imbalanced settings by learning reference-dependent representations from abundant reference data.
• Shape, Symmetries, and Structure: The Changing Role of Mathematics in Machine Learning Research — Mathematics' role in ML is shifting from theoretical guarantees to post-hoc explanations and high-level architectural guidance, leveraging advanced pure math.
• Probabilistic ML - 23 - Variational Inference — Variational inference approximates intractable posteriors by iteratively maximizing the ELBO within a tractable family of distributions.
• Quantifying and Understanding Uncertainty in Large Reasoning Models — A new method quantifies LRM uncertainty with statistical guarantees, linking reasoning traces to final answers.
• Mathematics for Data Science #5: Probability Distributions (The Real World Behaves This Way) — Understanding specific probability distributions is crucial for accurately modeling real-world data in data science.
• Generative models for decision-making under distributional shift — Generative models transform nominal data distributions into decision-relevant uncertainty distributions for OR problems.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Takara TLDR - Daily AI Papers.