传统信息量法将所有评价因子的信息量值进行综合，未对评价因子作优化选择，选择带有主观性。鉴于此，采用结果验证对信息量法进行改进，选取工程岩组、断层、河流、道路、坡度、坡向、剖面曲率7个评价因子。首先利用地质灾害编录数据与评价因子进行空间叠加分析，计算单因子信息量，然后利用成功率曲线法评价单因子评价结果的模型拟合能力，按照模型拟合能力从高到低将单因子信息量排序，并依次进行总信息量试算，得到7组试算结果。最后对7组试算结果进行模型拟合能力评价，确定AUC为前4的因子组合为最优的因子组合。%The traditional information quantity model synthesizes all the evaluation factors without optimization, which is sub-jective. Therefore, the results validation is adopted to improve the information quantity model by selecting seven evaluation fac-tors, namely lithology, fault, stream, road, slope ratio, slope direction and profile curvature. Firstly, the geological logging data and the evaluation factors are employed to carry out an overlay analysis, so the information value of single factor is obtained. Sec-ondly, the success ratio curve method is used to evaluate the model fitting capacity of the single factor;the information quantity of evaluation factors are ranked according to the model fitting capacity, and seven sets of total information quantity are obtained by trial calculation. Finally, the calculation results are evaluated so as to provide an assessment to the model fitting capacity, and the combination of top 4 AUC factors is determined as the best.
展开▼
