A random set and rule-based regression model incorporating image labels

Authors:
Guanyi Li;Jonathan Lawry
Affiliations:
Intelligent System Laboratory, Department of Engineering Mathematics, University of Bristol, Bristol, UK;Intelligent System Laboratory, Department of Engineering Mathematics, University of Bristol, Bristol, UK
Venue:
IUKM'13 Proceedings of the 2013 international conference on Integrated Uncertainty in Knowledge Modelling and Decision Making
Year:
2013

Citing 9
Cited 0

The transferable belief model

Artificial Intelligence
The nature of statistical learning theory

The nature of statistical learning theory
A framework for linguistic modelling

Artificial Intelligence
Application of fuzzy Naive Bayes and a real-valued genetic algorithm in identification of fuzzy model

Information Sciences—Informatics and Computer Science: An International Journal
Modelling and Reasoning with Vague Concepts (Studies in Computational Intelligence)

Modelling and Reasoning with Vague Concepts (Studies in Computational Intelligence)
Uncertainty modelling for vague concepts: A prototype theory approach

Artificial Intelligence
LIBSVM: A library for support vector machines

ACM Transactions on Intelligent Systems and Technology (TIST)
Genetic fuzzy learning

IEEE Transactions on Evolutionary Computation
Forecasting time series with genetic fuzzy predictor ensemble

IEEE Transactions on Fuzzy Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

A new form of conditional rules is proposed for regression problems in which a rule associates an input label with a corresponding image label on the output space. Here input labels are interpreted in terms of random set and prototype theory, so that each label is defined by a random set neighbourhood around a prototypical value. Within this framework we propose a rule learning algorithm and test its effectiveness on a number of benchmark regression data sets. Accuracy is compared with other several state-of-the-art regression algorithms, suggesting that our approach has the potential to be an effective rule learning methodology.