Search Swinburne Research Bank
Please use this identifier to cite or link to this item: http://hdl.handle.net/1959.3/3664
- A comparison of neural network input vector selection techniques
- Choi, Belinda; Hendtlass, Tim; Bluff, Kevin
- One of the difficulties of using Artificial Neural Networks (ANNs) to estimate atmospheric temperature is the large number of potential input variables available. In this study, four different feature extraction methods were used to reduce the input vector to train four networks to estimate temperature at different atmospheric levels. The four techniques used were: genetic algorithms (GA), coefficient of determination (CoD), mutual information (MI) and simple neural analysis (SNA). The results demonstrate that of the four methods used for this data set, mutual information and simple neural analysis can generate networks that have a smaller input parameter set, while still maintaining a high degree of accuracy.
- Publication type
- Conference paper
- Research centre
- Swinburne University of Technology. School of Information Technology
- Lecture notes in computer science : innovations in applied artificial intelligence : Proceedings of the 17th International Conference on Industrial and Engineering Applications of Artificial Intelligence and Expert Systems (IEA/AIE 2004), Ottawa, Canada, 17-20 May 2004, Vol. 3029 (2004), pp. 1-10
- Publication year
- Artificial neural networks; Coefficient of determination; Genetic algorithm; Mutual information
- Publisher URL
- Copyright © Springer-Verlag Berlin Heidelberg 2004.
- Peer reviewed