Daily reservoir inflow forecasting using artificial neural networks with stopped training approach

In this paper, an early stopped training approach (STA) is introduced to tr ain multi-layer feed-forward neural networks (FNN) for real-time reservoir inflow forecasting. The proposed method takes advantage of both Levenberg-M arquardt Backpropagation (LMBP) and cross-validation technique to avoid und erfitting or overfitting on FNN training and enhances generalization perfor mance. The methodology is assessed using multivariate hydrological time ser ies from Chute-du-Diable hydrosystem in northern Quebec (Canada). The perfo rmance of the model is compared to benchmarks from a statistical model and an operational conceptual model. Since the ultimate goal concerns the real- time forecast accuracy, overall the results show that the proposed method i s effective for improving prediction accuracy. Moreover it offers an altern ative when dynamic adaptive forecasting is desired. (C) 2000 Elsevier Scien ce B.V. All rights reserved.