摘要
增量神经网络(IncNet)的结构是由增长和剪枝控制,并且与训练数据的复杂性相匹配。用扩展卡尔曼滤波算法作为其学习算法。双径向转移函数比其它常用于人工神经网络的转移函数更具有灵活性。最新的改进是在多维空间中(具有N-1个参数)增加转移函数的旋转常数值。通过对逼近基准和心理分类问题的结果分析,清楚地表明比其他分类网络模型具有更强的泛化性。
Structure of incremental neural network （IncNet）is controlled by growing and pruning to match the complexity of training data. Extended Kalman Filter algorithm used as learning algorithm. Bi-radial transfer functions, more flexible than other functions commonly used in artificial neural networks. The latest improvement added is the ability to rotate the contours of constant values of transfer functions in multidimensional spaces with only N-1 adaptive parameters. Results on approximation benchmarks and on the real world psychometric classification problem clearly show superior generalization performance of presented network comparing with other classification models.
出处
《价值工程》
2013年第2期303-304,共2页
Value Engineering
关键词
人工神经网络
逼近
分类
转移函数
径向基函数
artificial neural flat work
approximation
classification
transfer function
RBF
