【摘要】 该文对应用知识发现技术训练神经元网络集成的方法进行了研究,提出了以并行操作的方式结合归纳学习所获取的知识和演绎学习所获取的知识的神经元网络集成模型KBNNE(Knowledge-basedNeuralNetworkEnsem-bles)。实验表明,通过调节所获取知识的权重因子,新模型可以有效提高网络集成的性能。更多还原

【Abstract】 This article explores the utility of knowledge discovery in Neural Network Ensembles. A novel neural network ensemble model KBNNE(Knowledge-Based Neural Network Ensembles)integrating KDD(Knowledge Discovery in Database)techniques and neural network ensemble algorithms by parallel operations is proposed. By balancing the relative importance of knowledge learned by induction and deduction, the new model improves the quality of neural network ensembles and has been applied successfully to actual modeling problems.更多还原