Associations System for Breast Cancer Microarray Data

The aim of this paper is to give biologists a tool to explore reasons and impacts of the breast cancer as patients with the same stage of illness can have different treatment responses. This paper proposes a Breast Cancer Associations system (BCA) to discover and interpret the associations among the breast cancer patient’s gene expressions data. The data used in this paper is the array data of 24.483 gene expression measurements recorded for 19 breast cancer patients. BCA consists of: data preprocessing, and data mining. In the first process in BCA, the data is carried out four preprocessing steps to be suitable and enhance the second process in BCA. These four steps are data filtration, normalization, discretization, and data adaptation. The mining process stage uses a new algorithm called Row Intersection Support Starting (RISS), which traverse the row enumeration space using the user-defined mines up threshold as a starting point deploying a new data format called Row Set (RS). The last stage in the system concerns the production of the association rules based on the user defined minimum confidence threshold. Fifteen different experiments have been conducted with different parameters. The results of the experiments are recorded and compared.