BP神经网络在器乐演奏评价中的应用

Application of Back Propagation Neural Network in the Instrumental Performance Assessment System

评价者在为器乐演奏者的表演打分时易受主观因素影响,而演奏者的表演也需从多个指标来衡量。BP神经网络可模拟由各因素构成的非线性映射。建立声乐评价体系,将各个评价指标客观化,使用成绩向量作为输入,经由BP神经网络输出得到合理的分数。实验仿真表明,通过训练的BP网络可模拟一个稳定的评分系统。

The evaluators are easy impacted by subjective factors when giving a correct assessent for instrumental performers＇ performace,and the performance also need to be measured from several indicators.BP neural network can simulate a nonlinear system.Establishes vocal evaluation model to make all indicators objective.The model uses score vector as an input.BP neural network will output an reasonable score.The result of simulation shows that a trained BP neural network can simulate a steady scoring system.