Feature Selection of Combining Relieff and Rough Set for Syndrome Classification of Chronic Gastritis in Traditional Chinese Medicine

Typically, the main form of Chinese medicine is based on interrogation of the patient's condition and artificial judgement by asking related patient symptom information; which is strongly subjective and prone to errors of judgment, leading to the wrong treatment outcome. The development in computer greatly improved the level of research in medicine, and also allowed the gradually objective and systematic improvement of experience knowledge. By analyzing TCM inquiry chronic gastritis data, the Relieff & Rough Set feature selection method was presented by combining different classification algorithms with experiments and analysis; the experimental results showed that efficient feature selection method can greatly enhance the effect of the classification; therefore, Relieff & Rough Set can be used as an efficient tool for feature selection and applied in the syndrome classification of TCM.