在不确定数据集上挖掘优化的概率干预策略

Mining Optimized Probabilistic Intervention Strategy over Uncertain Data Set

提出了不确定干预分析模型,主要工作包括:(1)建立了用于多维不确定数据分析的不确定监测点模型(uncertain surveillance);(2)建立了基于不确定监测点的不确定干预策略及挖掘评价算法;(3)在真实数据及仿真数据上对所提出的两种算法作了大量实验比较,验证了所提出的干预策略评价优化算法具有较高精度,效率比朴素方法高出3个数量级,适合在实际系统中处理海量干预评价.

This study provides models to analyze the intervention over uncertain data. The main contributions include： （1） It establishes a base model named Uncertain Surveillance; （2） It proposes probabilistic intervention strategies over uncertain surveillance with evaluating algorithms; （3） It gives extensive experiments and makes comparisons to show that the proposed model is highly precise and efficient in three order magnitudes of naive methods.