为提高中医诊断的智能化以及辩证的准确度, 提出一种基于多模态扰动策略的集成学习算法 (MPEL算法) .首先, 在样本域多次抽样产生不同的样本子空间;其次, 在属性域采用改进的层次聚类特征选择算法, 划分不同的属性子空间, 进而训练出具有较大差异性的基分类器;然后, 采用贪心策略选取最优的基分类器组合, 提高算法整体性能.选择中医哮喘病症状-证型病案进行验证, 并与其它集成学习算法对比, 实验结果表明, 改进的集成学习算法在哮喘病症状-证型分类预测中训练速度较快、识别准确率更高, 最高识别率高达98.16％.%In order to improve the intelligence and dialectical accuracy of the diagnosis of Chinese medicine, we propose an ensemble learning algorithm based on multi-modal perturbation strategy, called MPEL algorithm. Firstly, different sample sub-spaces of the sample domain are generated by multiple sampling. Secondly, different attribute sub-spaces are separated by an improved hierarchical clustering feature selection algorithm in the attribute space, and base classifiers with great diversity are trained. Thirdly, the optimal combination of base classifiers is selected through the greedy strategy so that the overall performance of the algorithm is improved. The medical records of Chinese medicine asthma symptoms-syndromes are selected to verify the performance of the proposed algorithm. Experimental results show that the proposed algorithm has faster training speed and higher recognition accuracy in the prediction of asthma symptoms-syndromes than other current ensemble learning algorithms, and the highest recognition accuracy of the MPEL can reach up to 98.16％.
展开▼
