Potential Risk Factors and Association of Significant Factors of Blood Cancer in Bangladesh Using Data Mining Techniques

Abstract

Cancer is now one of the leading causes of death globally and in Bangladesh. This research is focused on the blood cancer patients because there are only a few studies in Bangladesh on blood cancer. Initially, we collected 340 data (blood cancer and non-blood cancer) from BSMMU hospital. Then we use data mining to rank 30 factors associated with blood cancer. Our data mining methods include classification, chi-square (χ2) test, P-value, and association rule mining. The most potent predictors (P-value < .001) were muscle pull, inability to control the bladder, unusual bleeding, fever/raised temperature, and weakness in the legs. To predict an association between these significant elements, popular rule mining algorithms such as Apriori or Tertius are used. The results of the experiment show that weakness in the legs, fever and inability to control the bladder are common rules of blood cancer. Again, unusual bleeding, rapid illness, and muscle pull are likely to be associated. A fair skin tone with rapid breathing and leg weakness may also be a danger. Journal of Engineering Science 13(1), 2022, 9-20