Association Rule Discovery with Fuzzy Decreasing Support on Syndrome Differentiation in Coronary Heart Disease

Association rules represent a promising technique to search syndrome differentiation on modern Chinese medicine. Over the years, a variety of algorithms for finding frequent itemsets in very large transaction databases have been developed. The key feature in most of these algorithms is that they use a constant support constraint to control the inherently exponential complexity of problem. Long itemsets with low support can still be interesting but it is unable to find them. This paper presents a new association rules mining framework: fuzzy decreasing support-confidence to find all itemsets that satisfy a length-decreasing support constraint. We extract data about relevant factors of syndrome differentiation from the coronary heart disease data collected from hospital. The experimental results show that the frameworks proposed in this paper can not only verify the existing syndrome differentiation, but also can discover syndrome differentiation with a combination of multiple factors.