Continuous Analogical Proportions-Based Classifier
- 5 June 2020
- book chapter
- Published by Springer Nature
- Vol. 1237, 541-555
Abstract
Analogical proportions, often denoted \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$A:B\,{::}\,C:D$$\end{document} , are statements of the form “A is to B as C is to D” that involve comparisons between items. They are at the basis of an inference mechanism that has been recognized as a suitable tool for classification and has led to a variety of analogical classifiers in the last decade. Given an object D to be classified, the basic idea of such classifiers is to look for triples of examples (A, B, C), in the learning set, that form an analogical proportion with D, on a maximum set of attributes. In the context of classification, objects A, B, C and D are assumed to be represented by vectors of feature values. Analogical inference relies on the fact that if a proportion \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$A:B\,{::}\,C:D$$\end{document} is valid, one of the four components of the proportion can be computed from the three others. Based on this principle, analogical classifiers have a cubic complexity due to the search for all possible triples in a learning set to make a single prediction. A special case of analogical proportions involving only three items A, B and C are called continuous analogical proportions and are of the form “A is to B as B is to C” (hence denoted \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$A:B\,{::}\,B:C$$\end{document} ). In this paper, we develop a new classification algorithm based on continuous analogical proportions and applied to numerical features. Focusing on pairs rather than triples, the proposed classifier enables us to compute an unknown midpoint item B given a pair of items (A, C). Experimental results of such classifier show an efficiency close to the previous analogy-based classifier while maintaining a reduced quadratic complexity.
Keywords
This publication has 9 references indexed in Scilit:
- An Analogical Interpolation Method for Enlarging a Training DatasetLecture Notes in Computer Science, 2019
- Making the Best of Cases by Approximation, Interpolation and Extrapolation2018
- Analogical proportions: From equality to inequalityInternational Journal of Approximate Reasoning, 2018
- Oddness-based classification: A new way of exploiting neighborsInternational Journal of Intelligent Systems, 2018
- Analogy-based classifiers for nominal or numerical dataInternational Journal of Approximate Reasoning, 2017
- Multiple-valued extensions of analogical proportionsFuzzy Sets and Systems, 2016
- Analogical proportions: another logical view2009
- Handling Analogical Proportions in Classical Logic and Fuzzy Logics Settings2009
- Analogical Dissimilarity: Definition, Algorithms and Two Experiments in Machine LearningJournal of Artificial Intelligence Research, 2008