Ensemble modelling or selecting the best model: Many could be better than one

Authors:
S. V. Barai;Yoram Reich
Affiliations:
Department of Civil Engineering, Indian Institute of Technology, Kharagpur-721 302, West Bengal, India;Department of Solid Mechanics, Materials and Structures, Faculty of Engineering, Tel Aviv University, Ramat Aviv 69978, Israel
Venue:
Artificial Intelligence for Engineering Design, Analysis and Manufacturing
Year:
1999

Citing 9
Cited 5

What size net gives valid generalization?

Neural Computation
Neurocomputing

Neurocomputing
The gap between abstract and concrete results in machine learning

Journal of Experimental & Theoretical Artificial Intelligence
Neural networks and the bias/variance dilemma

Neural Computation
Original Contribution: Stacked generalization

Neural Networks
Machine learning, neural and statistical classification

Machine learning, neural and statistical classification
Universal approximation using feedforward neural networks: a survey of some existing methods, and some new results

Neural Networks
Neural Networks for Pattern Recognition

Neural Networks for Pattern Recognition
Learning in design: From characterizing dimensions to working systems

Artificial Intelligence for Engineering Design, Analysis and Manufacturing

2007 Special Issue: Reducing uncertainties in neural network Jacobians and improving accuracy of neural network emulations with NN ensemble approaches

Neural Networks
Soft computing in engineering design - A review

Advanced Engineering Informatics
Development of a parsimonious GA-NN ensemble model with a case study for Charpy impact energy prediction

Advances in Engineering Software
Bankruptcy prediction using Extreme Learning Machine and financial expertise

Neurocomputing
Ensemble delta test-extreme learning machine (DT-ELM) for regression

Neurocomputing

Quantified Score

Hi-index	0.00

Visualization

Abstract

In the course of data modelling, many models could be created. Much work has been done on formulating guidelines for model selection. However, by and large, these guidelines are conservative or too specific. Instead of using general guidelines, models could be selected for a particular task based on statistical tests. When selecting one model, others are discarded. Instead of losing potential sources of information, models could be combined to yield better performance. We review the basics of model selection and combination and discuss their differences. Two examples of opportunistic and principled combinations are presented. The first demonstrates that mediocre quality models could be combined to yield significantly better performance. The latter is the main contribution of the paper; it describes and illustrates a novel heuristic approach called the SG(k-NN) ensemble for the generation of good-quality and diverse models that can even improve excellent quality models.