Abstract: A dynamic forecasting system of demands for farm products in consuming market is studied by means of applying the intelligent forecast method SVM(Support Vector Machine).In order to achieve higher forecast precision of farm product sale quantities,such factors as weather,climate conditions and demands on holidays etc.,are introduced to the forecasting model,and the Fuzzy Theory is also applied to cope with the problem of fuzzification.Furthermore,an intelligent forecasting system is constructed,which is mainly based on SVM theory and combined with many other techniques.Using this system to dynamically forecast the demands for farm products,the practical application shows the precision of the forecasting method.

Key words: support vector machine, farm products, sales forecast, fuzzy theory, kernel