多重改进型指数双向联想记忆模型及其在多证据推理中的决策性能

提出了多证据推理中采用神经网络来模拟信念组合学习方法。网络由多个改进型指数双向联想记忆模型（IeBAM）构成，并且共享一个输出来同时进行多证据不确定性的管理。文中证明了多重IeBAM(Multi-IeBAM）的稳定性，讨论了在多条证据同时提交网络后的多数规则。理论和实验都证明了多数因子比Wang所提模型更紧凑、更严格，从而可保证在受一定程度的干扰下，专家们仍能做出正确决策。最后所给出的模拟例子的结果与直觉推理相吻合。

Multiple Improved Exponential Bidirectional Associative Memory and Its Decision-Making Performance in Multievidence Reasoning

In this paper, a method for modeling the learning of belief combination in evidential reasoning using a neural network is presented. A centralized network composed of multiple improved exponential bidirectional associative memories (Multi-IeBAM) sharing a single output array of neurons is proposed to process the uncertainty management of many pieces of evidence simutaneously. The stability of the proposed Multi-IeBAM network is proved. A majority rule of decision making in presentation of multiple evidence is also found by the study of signal-noise-ratio of multiple IeBAM networks. The theoretical analysis and computer s imulations indicate that the majority factor of Multi-IeBAM is tighter and stri cter than that of Multi-MeBAM and Multi-eBAM, which grants that experts co uld still reach the right decision under a certain degree of disturbance in evid ence. The result of simulation analysis given in the end is coincident with the intuition of reasoning.