The Duobaoshan ore concentration area is located in the northern part of Heilongjiang Province, China, and stands as the largest polymetallic mining area in Northeast China. Currently, deposits such as porphyritic copper (molybdenum) deposits, epithermal gold deposits, and skarn-type iron-copper deposits have been discovered in this region, indicating significant metallogenic potential. The combination of machine learning techniques and multivariate geoscience data for quantitative metallogenic prediction has been a focal point of research in geology and resource exploration in recent years. Taking the Duobaoshan mining area as an example, this article constructs a mineral prediction process based on machine learning methods: (1) Establish a metallogenic model and identify ore-controlling factors through thorough research on the metallogenic system in this area; (2) Develop a prospecting model grounded in comprehensive ore-controlling factors and prospecting indicators to provide pertinent exploration data for evaluation and prediction; (3) Informatize the prospecting model and establish a prediction model, extracting seven layers of prediction factors in total; (4) Utilize three distinct machine learning methods to quantitatively predict the prospect areas for copper-molybdenum mineralization in the study area. The prediction results indicate that the Duobaoshan mining area holds promising prospecting potential, with two Grade I target areas and two Grade II target areas delineated, thus providing a scientific basis for further exploration and prospecting endeavors.
Read full abstract