Integration of neural networks and expert systems for time series prediction
Ventsislav Georgiev Nikolov, Valeri Bogdanov · 2010
In this paper an approach for development of univariate time series prediction library in Java and its integration with an existing CLIPS related system is presented. A backpropagation neural network based forecaster is provided in the current version of the library. The connection between Java and CLIPS is realized by Java Native Interface (JNI) and CLIPS user defined functions. The expert system is used to construct the object structure, configure and fit the prediction model, thus reducing user expertize requirements and allowing for complete automation.