A multi-instance multi-label learning approach to objective auscultation analysis of traditional Chinese medicine

The purpose of this paper is to study objective auscultation of traditional Chinese medicine using multi-instance multi-label (MIML) learning. The experiment data are the patients' speech samples of 5 vowels i.e. /a/,/o/,/e/,/i/,/u/. Each patient in the dataset may have one or both of the qi and yin deficiency syndromes. By regarding the 5 vowel samples from one patient as instances and the patient's syndrome type as the labels, the problem can be properly formalized under multi-instance multi-learning framework. In the conducted experiment, features are extracted from the speech samples and processed by MIML algorithm for classification. Satisfactory performance is obtained which proves that MIML is an effective and feasible approach for auscultation analysis.