基于主成分分析与神经网络的采矿方法优选
来源期刊:中南大学学报(自然科学版)2010年第5期
论文作者:陈建宏 刘浪 周智勇 永学艳
文章页码:1967 - 1972
关键词:采矿方法;主成分分析法;BP神经网络
Key words:mining methods; principal components analysis; BP-neural networks
摘 要:基于利用神经网络预测采矿方法存在一些不足,建立主成分分析法与神经网络结合的采矿方法优选模型。对神经网络的输入数据进行主成分分析,使输入数据不相关且减少。研究结果表明:利用主成分分析法可将输入数据减少,消除由于BP网络输入数据太多而影响数据处理速度的缺陷;把主成分分析法和神经网络结合进行采矿方法优选,可使预测精度大大提高。
Abstract: Based on the fact that the deficiencies of neural networks predict mining method, the optimization model of mining methods was set up combining principal component analysis with neural networks. The results show that using the principal component analysis can reduce input-data, eliminate the defect which is due to BP network input data too much influencing the data-processing speed. Combining principal components analysis with neural networks to optimize mining method can make the prediction accuracy improve greatly.