Discovering patterns of NED-breast cancer based on association rules using apriori and FP-growth

No Evidence of Disease (NED) is breast cancer patient condition status which it indicates that they can life, no find the cancer by tested, and without any symptoms of cancer in period of times, after they received primary treatment. NED is a critical status, because it involves the treatment type and patient cancer condition factors. This paper examines about breast cancer problem in data mining technical side, especially to discover the patterns of NED-breast cancer patient using cancer registry data from Oncology Hospital. Its patterns are discovered through the relationship of among features begin from 1dimensional, 2-dimensional, 3-dimensional, and n-dimensional. We applied association rules mining using Apriori and FP-Growth algorithm, which both have the advantage and drawback. Apriori algorithm involves all generation of candidate item sets and multiple database scans, but it makes highconsuming iteration. While FP-Growth algorithm extracts the frequent item sets directly from FP-Tree, it make the advantage of FP-Growth that is faster process needs only scan the database once. This paper experiment shown that the association result of Apriori and FP-Growth is almost similar, 10-highest confidence value represented 100% confidence of association rule on breast cancer dataset with support value up to 50%.