论文部分内容阅读
运用人工神经网法对复合氟化物（ＡＢｍＦｎ）中Ｅｕ＋离子光谱结构进行了分类研究，对神经网络的基本特征进行了观察，实验表明，神经网的结构、初始权重以及作为训练集和预测集的随机分组对最终结果都会产生重要影响。本研究采用测试集来监控训练过程，以避免“过训练”，并提高神经网的预测性能。方法识别率达到１００％，预测率达到９６．３％．结果明显优于常规模式识别法。
The classification of Eu + ions in the complex fluoride (ABmFn) was studied by using artificial neural network. The basic characteristics of the neural network were observed. The experimental results show that the structure of the neural network, the initial weights and the training set and prediction set Random grouping will have a significant impact on the final result. This study uses a test suite to monitor the training process to avoid “over training” and to improve neural network predictive performance. The recognition rate of the method reached 100% and the prediction rate reached 96.3%. The result is obviously better than the conventional pattern recognition method.