A meta-learning approach to select meta-heuristics for the traveling salesman problem using MLP-Based label ranking

Authors:
Jorge Kanda;Carlos Soares;Eduardo Hruschka;Andre de Carvalho
Affiliations:
Instituto de Ciencias Matematicas e de Computacao, Universidade de Sao Paulo, Sao Carlos, SP, Brazil,Instituto de Ciencias Exatas e Tecnologias, Universidade Federal do Amazonas, Itacoatiara, AM, ...;INESC TEC Porto LA/Faculdade de Economia, Universidade do Porto, Porto, Portugal;Instituto de Ciencias Matematicas e de Computacao, Universidade de Sao Paulo, Sao Carlos, SP, Brazil;Instituto de Ciencias Matematicas e de Computacao, Universidade de Sao Paulo, Sao Carlos, SP, Brazil
Venue:
ICONIP'12 Proceedings of the 19th international conference on Neural Information Processing - Volume Part III
Year:
2012

Citing 16
Cited 0

Learning internal representations by error propagation

Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
A user's guide to tabu search

Annals of Operations Research - Special issue on Tabu search
A perspective view and survey of meta-learning

Artificial Intelligence Review
Ranking Learning Algorithms: Using IBL and Meta-Learning on Accuracy and Time Results

Machine Learning
AllelesLociand the Traveling Salesman Problem

Proceedings of the 1st International Conference on Genetic Algorithms
Introduction to Data Mining, (First Edition)

Introduction to Data Mining, (First Edition)
The Traveling Salesman Problem: A Computational Study (Princeton Series in Applied Mathematics)

The Traveling Salesman Problem: A Computational Study (Princeton Series in Applied Mathematics)
Multilabel classification via calibrated label ranking

Machine Learning
The WEKA data mining software: an update

ACM SIGKDD Explorations Newsletter
Metalearning: Applications to Data Mining

Metalearning: Applications to Data Mining
Understanding TSP difficulty by learning from evolved instances

LION'10 Proceedings of the 4th international conference on Learning and intelligent optimization
Handbook of Metaheuristics

Handbook of Metaheuristics
Selection of algorithms to solve traveling salesman problems using meta-learning

International Journal of Hybrid Intelligent Systems - Feature and algorithm selection with Hybrid Intelligent Techniques
Review: Measuring instance difficulty for combinatorial optimization problems

Computers and Operations Research
Ant colony system: a cooperative learning approach to the traveling salesman problem

IEEE Transactions on Evolutionary Computation
No free lunch theorems for optimization

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Different meta-heuristics (MHs) may find the best solutions for different traveling salesman problem (TSP) instances. The a priori selection of the best MH for a given instance is a difficult task. We address this task by using a meta-learning based approach, which ranks different MHs according to their expected performance. Our approach uses Multilayer Perceptrons (MLPs) for label ranking. It is tested on two different TSP scenarios, namely: re-visiting customers and visiting prospects. The experimental results show that: 1) MLPs can accurately predict MH rankings for TSP, 2) better TSP solutions can be obtained from a label ranking compared to multilabel classification approach, and 3) it is important to consider different TSP application scenarios when using meta-learning for MH selection.