Identifying characteristics of seaports for environmental benchmarks based on meta-learning

Authors:
Ana Ximena Halabi Echeverry;Deborah Richards;Ayse Bilgin
Affiliations:
Department of Computing, Macquarie University, NSW, Australia;Department of Computing, Macquarie University, NSW, Australia;Department of Statistics, Macquarie University, NSW, Australia
Venue:
PKAW'12 Proceedings of the 12th Pacific Rim conference on Knowledge Management and Acquisition for Intelligent Systems
Year:
2012

Citing 4
Cited 0

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

Machine Learning
Distributed prediction from vertically partitioned data

Journal of Parallel and Distributed Computing
Characteristics analysis for small data set learning and the comparison of classification methods

AIKED'08 Proceedings of the 7th WSEAS International Conference on Artificial intelligence, knowledge engineering and data bases
UCI++: Improved Support for Algorithm Selection Using Datasetoids

PAKDD '09 Proceedings of the 13th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we discuss a model which classifies any seaport in the context of environmental management system standards as leader, follower and average user. Identification of this status can assist Port Authorities (PAs) in making decisions concerned with finding collaborating seaport partners using clear environmental benchmarks. This paper demonstrates the suitability of meta-learning for small datasets to assist pre-selection of base-algorithms and automatic parameterization. The method is suitable for small number of observations with many attributes closely related with potential issues concerning environmental management programs on seaports. The variables in our dataset cover main aspects such as reducing air emissions, improving water quality and minimizing impacts of growth. We consider this model will be suitable for Port authorities (PAs) interested in effective and efficient methods of knowledge discovery to be able to gain the maximum advantage of benchmarking processes within partner ports. As well as for practitioners and non-expert users who want to construct a reliable classification process and reduce the evaluation time of data processing for environmental benchmarking.