A hybrid architecture for predicting oil slick movement

Abstract

In this paper, we present a hybrid architecture for an intelligent system that can be used to project oil slick movement. The system under construction has the ability to learn from historical weather data and then to incorporate the learned knowledge into its projection of the future movement of oil slick. It employs probabilistic reasoning to deal with uncertainty in the observed data and weather forecast, neural networks to acquire knowledge from historical data and fuzzy logic to deal with imprecision embedded in the available information. This innovative approach to this highly complicated, but very important and practical issue exemplifies the application of advanced AI techniques to the practical problems.<>