铜锍吹炼终点预报系统样本的预处理

The Pretreatment of Samples of Copper - matte Converting Endpoint Prediction System

在进行网络学习前,对样本进行标准化、变量空间的降维、过滤噪音等处理,使得样本具有很强 的特征性,可提高网络的学习能力。

Before network study, the samples are made to have very strong characteristics by standardizing the samples, lowering the dimension of sample variable space, and filtering sample noise etc. These can improve the study ability of netvvork.