基于RUSBoost和积矩系数的神经网络分类算法

针对单个神经网络分类准确率低、RUSBoost算法提高NN分类器准确率耗时长的问题，提出了一种混合RUSBoost算法和积矩系数的分类优化算法。首先，利用RUSBoost算法生成m组训练集；然后，依据Pearson积矩系数计算每组训练集属性的相关程度消除冗余属性，生成目标训练集；最后，新的子训练集训练神经网络分类器，选择最大准确率分类器作为最终的分类模型。实验中使用了4个Benchmark数据集来验证本文算法的有效性。实验结果表明，本文提出的算法的准确率相较于传统的算法最大提升了8.26%，训练时间最高降低了62.27%。

An optimization algorithm for Neural network based on RUSBoost and correlation coefficient

Referring to the current problems that a single neural network classification accuracy was low, and RUSBoost RUSBoostis a time-consuming algorithm for improving the NN classifier accuracy. In this paper, we proposed a hybrid algorithm which combined RUSBoost RUSBoostwith the Pearson correlation coefficient to optimize the classifier of neural network. First of all, ?? is generated by using RUSBoost algorithm for group training; Then, according to the Pearson product-moment coefficient of correlation calculate the property of each group and eliminates redundant properties in the set, generating target training sets; Finally, neural network classifiers were trained by the new training set, and select the maximum accuracy classifier as the ultimate model. We choose 4 Benchmark data sets to verify the effectiveness of the algorithm, experimental results show that the accuracy of the algorithms presented compared to traditional algorithm for maximum lifting 8.26%, and the training timer reduced 62.27%.