摘要
本文提出矿物分类和识别的人工神经网络模型，并选取一组标样──我国沉积碳酸盐型锰矿中菱锰矿作为研究对象，识别效率达100％。结果表明，该模型性能良好，可望成为矿物识别的有效手段。
Presented in this paper is the artificial neural network model for classification and recognitionof minerals, and a group of samples of rhodochrosite from sedimentary carbonate-type manganese ores in China were collected as the subject of study. The successful rate reached 100%. Theresults show that the performance of the artificial neural network approach is good, and thereforeit may become an effective approach to the recognition of minerals.
出处
《矿物学报》
CAS
CSCD
北大核心
1994年第1期56-60,共5页
Acta Mineralogica Sinica
关键词
矿物
分类
神经网络
B-P模型
classification of mineral
artificial neural network
back-propagation model
