Evaluating the Effect of Different Mode's Attributes on the Subjective Classification in the Case of TCM

This paper proposes a method for assessing the subjective classifications of traditional Chinese medicine (TCM) and investigating the influence of attributes on them, while these attributes are extracted from multi-sensors and represented by different modes. In TCM, a person's health states can be represented by 13 Zhengs that are not entirely independent, while the diagnosis data given by TCM doctors are subjective. Accordingly, the influence of the modes and the attributes extracted from the multimodal sensor data on the Zheng's classification is validated by a defined aggregation function called aas. Moreover, the conditions of removing the weak modes are proposed based on the correlation between the attributes of modes and the number of the attributes close to the Zhengs. The simulation results verify the adequacy of the above aas and conditions in evaluating the effect of attributes on the classification performance.