Cancer Tumor Detection by Gene Expression Data Exploration Using a Genetic Fuzzy System

Classification of different tumors in cancer detection and drug identification is very important task. Prior cancer classification was based on clinical information that has a limited ability for detection and debugging. Recently, micro array technology has enabled monitoring the description of thousands of genes simultaneously. Rule-based expert systems are often used for decision support in various fields such as error detection, biology and medicine. In some fields like medicine it is preferred to use classifiers that are not in form of black box (like neural network) because it helps users to understand the knowledge of classifier. Fuzzy rule based classifiers are suitable for this matter because they are simply interpretable and they haven't deterministic classifiers limitation. In this paper, we proposed cancer detection on Global Cancer Map dataset by creating fuzzy rule with genetic algorithm. We'll show that our approach is useful in cancer tumor detection based on the results.