基于自适应核密度的贝叶斯概率模型岩性识别方法研究
蔡泽园1,2,3(
), 鲁宝亮1,2,3(), 熊盛青4, 王万银1,2,3
), 鲁宝亮1,2,3(), 熊盛青4, 王万银1,2,3
Lithology identification based on Bayesian probability using adaptive kernel density
Ze-Yuan CAI1,2,3(), Bao-Liang LU1,2,3(), Sheng-Qing XIONG4, Wan-Yin WANG1,2,3
), Bao-Liang LU1,2,3(), Sheng-Qing XIONG4, Wan-Yin WANG1,2,3
图9. 869点训练样本分类结果
a~f分别表示对于869个训练样本点的传统高斯分类、固定带宽的核密度估计、自适应带宽的核密度估计的贝叶斯分类结果以及其对应的概率分布
Fig.9. Classification results of 435 training samples
Figures a~frespectively represent the Bayesian classification results, corresponding probability distribution map of the traditional Gaussian classification, fixed bandwidth kernel density estimation, adaptive bandwidth kernel density estimation for 869 training sample points
