Dong Yilin, Li Xinde, Liu Yihai
Key Laboratory of Measurement and Control of CSE, Ministry of Education, School of Automation, Southeast University, Nanjing, Jiangsu Province, China.
Jiangsu Automation Research Institute, Lianyungang, Jiangsu Province, China.
PLoS One. 2018 Jan 19;13(1):e0189703. doi: 10.1371/journal.pone.0189703. eCollection 2018.
In many applications involving epistemic uncertainties usually modeled by belief functions, it is often necessary to approximate general (non-Bayesian) basic belief assignments (BBAs) to subjective probabilities (called Bayesian BBAs). This necessity occurs if one needs to embed the fusion result in a system based on the probabilistic framework and Bayesian inference (e.g. tracking systems), or if one needs to make a decision in the decision making problems. In this paper, we present a new fast combination method, called modified rigid coarsening (MRC), to obtain the final Bayesian BBAs based on hierarchical decomposition (coarsening) of the frame of discernment. Regarding this method, focal elements with probabilities are coarsened efficiently to reduce computational complexity in the process of combination by using disagreement vector and a simple dichotomous approach. In order to prove the practicality of our approach, this new approach is applied to combine users' soft preferences in recommender systems (RSs). Additionally, in order to make a comprehensive performance comparison, the proportional conflict redistribution rule #6 (PCR6) is regarded as a baseline in a range of experiments. According to the results of experiments, MRC is more effective in accuracy of recommendations compared to original Rigid Coarsening (RC) method and comparable in computational time.
在许多涉及通常由信度函数建模的认知不确定性的应用中，常常需要将一般的（非贝叶斯）基本信度分配（BBA）近似为主观概率（称为贝叶斯BBA）。如果需要将融合结果嵌入基于概率框架和贝叶斯推理的系统（如跟踪系统）中，或者如果需要在决策问题中做出决策，就会出现这种需求。在本文中，我们提出了一种新的快速组合方法，称为改进的刚性粗化（MRC），以基于识别框架的层次分解（粗化）获得最终的贝叶斯BBA。对于该方法，通过使用不一致向量和简单的二分法，有效地对具有概率的聚焦元素进行粗化，以降低组合过程中的计算复杂度。为了证明我们方法的实用性，将这种新方法应用于在推荐系统（RS）中组合用户的软偏好。此外，为了进行全面的性能比较，在一系列实验中，将比例冲突再分配规则#6（PCR6）作为基线。根据实验结果，与原始的刚性粗化（RC）方法相比，MRC在推荐准确性方面更有效，并且在计算时间上相当。
