Testing the applicability of artificial intelligence techniques to the subject of erythemal ultraviolet solar radiation. Part two: an intelligent system based on multi-classifier technique

J Photochem Photobiol B. 2008 Mar 28;90(3):198-206. doi: 10.1016/j.jphotobiol.2007.12.001. Epub 2007 Dec 14.

Abstract

The problem we address here describes the on-going research effort that takes place to shed light on the applicability of using artificial intelligence techniques to predict the local noon erythemal UV irradiance in the plain areas of Egypt. In light of this fact, we use the bootstrap aggregating (bagging) algorithm to improve the prediction accuracy reported by a multi-layer perceptron (MLP) network. The results showed that, the overall prediction accuracy for the MLP network was only 80.9%. When bagging algorithm is used, the accuracy reached 94.8%; an improvement of about 13.9% was achieved. These improvements demonstrate the efficiency of the bagging procedure, and may be used as a promising tool at least for the plain areas of Egypt.

MeSH terms

  • Algorithms
  • Artificial Intelligence*
  • Egypt
  • Erythema / etiology*
  • Models, Theoretical
  • Neural Networks, Computer
  • Reproducibility of Results
  • Time Factors
  • Ultraviolet Rays / adverse effects*