Time series forecasting in a distributed environment
Paulo Cortez, Jose Manuel Machado, José Neves · Portuguese National Funding Agency for Science, Research and Technology (RCAAP Project by FCT) · 1996
The area of Time Series Forecasting (forecasting observations ordered in time) is object of attention, in particular due to its role in the process of decision making. Artificial Neural Networks} (ANNs) have been successfully used in pattern recognition, even in Time Series with a high degree of noise, but fail in the attempt to develop a fine architecture on their own Genetic Algorithms (GAs), on the other hand, proved to be one of the most promising arenas to elaborate on problem solving optimization strategies. Amalgamating both paradigms, one makes the denizens of a new computational world: ANNs and GAs. ANNs will assimilate the pattern series and better ANNs architectures will be attained by evolution.