摘要： 介绍了建立热轧带钢力学性能质量模型的数据挖掘过程。用普通神经网络建立起由工艺参数预测力学性能的质量模型，模型预测结果的5%命中率是0.508。提出了一种新的建模方法──逐层逼近法，并用它建立起质量模型，预测结果的5%命中率达到0.721，完全可以满足现实生产需要。

关键词: 数据挖掘, 人工神经网络, 力学性能

Abstract: The data mining process of establishing the quality model is introduced, that could predict the mechanical properties of hot-rolled steel strip with the technological parameter. The quality model whose hit ratio of 5% deviation reach 0.508 is established by applying the technology of basic artificial neural network. The quality model whose hit ratio of 5% deviation reach 0.721 is established by applying the technology of layer-of-layer impending, and this model could meet current industrial demand fully.

Key words: Data mining, Artificial neural network, Mechanical properties