(1. 中南大學(xué) 有色金屬成礦預(yù)測教育部重點(diǎn)實(shí)驗(yàn)室，長沙 410083；
2. 中南大學(xué) 地球科學(xué)與信息物理學(xué)院，長沙 410083；3. 山金西部地質(zhì)礦產(chǎn)勘查有限公司，西寧 810016)
摘 要: 針對目前成礦信息不對稱等的問題，提出組合作用域信息集成方法，并以桂西—滇東南地區(qū)錳礦為例，利用線性回歸分析進(jìn)行了資源量預(yù)測。通過成礦信息場分析方法，提取某些地質(zhì)類的成礦信息，并建立了桂西—滇東南地區(qū)錳礦成礦預(yù)測指標(biāo)集，同時(shí)考慮到模型區(qū)和預(yù)測區(qū)的找礦信息不對稱問題，采用組合作用域集成線性回歸分析，建立研究區(qū)錳礦資源量預(yù)測模型。實(shí)例分析表明，采用組合作用域信息集成預(yù)測模型能精細(xì)、有效地對研究區(qū)中的錳礦資源量進(jìn)行預(yù)測，其預(yù)測結(jié)果對進(jìn)一步找礦具有很好的指導(dǎo)意義，利用組合作用域信息集成方法可幫助解決成礦信息不對稱的問題。
關(guān)鍵字: 成礦信息；組合作用域；線性回歸；成礦信息不對稱
condition of metallogenic information asymmetry
（1. Key Laboratory of Metallogenic Prediction of Nonferrous Metals, Ministry of Education,
Central South University, Changsha 410083, China;
2. School of Geosciences and Info-Physics, Central South University, Changsha 410083, China;
3. Shanjin Western Geological and Mineral Exploration Co., Ltd., Xining 810016, China）
Abstract:Focusing on the issue of metallogenic information asymmetry, a new method of metallogenic information integration based on combination domains was proposed, and an instance study of manganese resources prediction using the linear regression analysis was made by taking the Western Guangxi—Southeastern Yunnan areas as an example. Through the field analysis method of metallogenic information, some kinds of geological metallogenic informations were extracted, and the index set of metallogenic prognosis in the West Guangxi—Southeast Yunnan was built. Considering the issue of metallogenic information asymmetry between the model area and the prediction area, the prediction model of manganese resource was built using the linear regression analysis based on the integration of combination domains. The instance study indicates that the linear regression model based on integration of combination domains can be used to predict the manganese resource more precisely and effectively, which can resolve some issues of metallogenic information asymmetry, and the prediction results have indicative significance to the further predicting.
Key words: metallogenic information; combination domains; linear regression; metallogenic information asymmetry
