Novel improved SMOTE resampling integrated algorithm based onfractal for geochemical anomalies evaluation
基于分形SMOTE重采样集成算法圈定区域化探异常

Based on the similarity theory of fractal, this paper put forward a new SMOTE re-sampling algorithm. According to the real distribution of samples, a few sets of data samples should be reconstructed to realize the equalization of data sets. The new algorithm combined Adaboost technology, according to the classification of the error rate updating weights of samples to improve the classification performance of imbalanced data. The new algorithm was based on the simulation experiment on the research objection of polymetallic deposits such as tin and copper from Gejiu, Yunnan province. The experimental results show that predicted results for the new algorithm delineating regional geochemical anomalies are better than traditional methods, which can identify the geochemical anomaly accurately.