基于聚类的神经网络规则抽取算法

为了从人工神经网络中抽取规则，提出一种新的规则抽取算法。网络被训练并剪枝后，将隐节点的激活值离散化，对输入到隐节点的权重进行聚类，聚类过程中可根据隐节点的激活值动态调整权值聚类数目，进而高效准确地抽取规则。实验结果表明,该算法可明显降低规则抽取的时间复杂度，减少生成规则的数量。

Rules Extraction from Artificial Neural Network Based on Clustering

We propose a novel algorithm for extracting rules from artificial neural network. After the network is trained and pruned successfully, the activation values at the hidden unit are clustered into discrete values. In the cluster phase, the cluster number of weights can be adjusted dynamically according to activation values of their corresponding hidden units. The experiment results have shown its feasibility and accuracy, and prove that this algorithm decreases the complexity of rules extraction and makes the number of the extracted rules small.