Intelligent clustering techniques for prediction of sugar production

The accurate, and timely prediction of the annual sugar-beet crop yield is important to Sugar Industry because, based on it, the “harvest campaign” can be scheduled efficiently. This work presents intelligent clustering techniques for effecting efficient, small error prediction of the annual sugar-beet crop yield for the Hellenic Sugar Industry based on production and meteorological data acquired during a period of 11 years. The experiments here demonstrate that intelligent clustering techniques can provide with better estimates of sugar production than alternative prediction methods including an “energy conservation” system model.