針對信息系統的特點和小波基具有很強的自適應性和函數變化能力，提出了一種基于小波神經網絡的信息系統綜合評價模型，并在最小均方能量準則下，采用相應的共軛梯度學習算法求解小波函數的尺度和時延參數，以及小波網絡的權值。仿真實驗表明該算法用于信息系統綜合評價的有效性與準確性，為信息系統的評價研究提供一條新的途徑。
關 鍵 詞 小波神經網絡; 信息系統; 綜合評價; 算法

Abstract This paper proposes a new comprehensive valuation method based on characteristic of information system and the property that wavelet can smooth data and adapt to the change of function. Furthermore, the dilation and shifts that are the parameters of wavelet function and the weight of neural network structure are calculated by conjugated gradient algorithm in term of the latest mean squared energy criterion.The results shown by the simulation that the method was effective and veracity for comprehensive evaluation of information system, and can provide a new way for evaluation research of information system.
Key words wavelet neural network; information system; comprehensive evaluation; algorithm