Study of Long-term Runoff Forecast Model Based on Association Rules Mining
Shiguo Xu · South-to-north Water Transfers and Water Science & Technology · 2007
Association rule is an important method of data mining techniques. Considering the characteristics of hydrology forecast, association rules method is applied to the long-term runoff forecast. The forecast factors are selected to constitute the long-term forecast database on the forecast object. The original data is discretized to find the strong association rules which accords with the minimum-support and minimum-confidence. Long-term runoff forecast model is built based on the strong association rules. According to the practical example of Jiangqiao hydrologic station at Nenjiang River, the strong association rules reveal the relation between the sea surface temperature (SST) and the flood season runoff, and the association rules method is proved feasible in aided decision-making for long-term runoff forecast.