Applying Computational Intelligence Techniques For Diagnoses of Breast Cancer

Abstract

Cancer, in particular breast cancer, is considered one of the most common causes of death worldwide, according to the world health organization. For this reason, extensive research efforts have been done in the area of accurate and early diagnosis of cancer in order to increase the likelihood of cure. Among the available tools for diagnosing cancer, microarray technology is commonly used in biological and medical science to study gene expression in cells. When a healthy tissue becomes cancerous, gene expression levels change. Tissues can be classified by looking for changes in gene expression. Although the huge number of features or genes in the microarray data may seem advantageous, many of these features are irrelevant or redundant, resulting in the deterioration of classification accuracy. To overcome this challenge, feature selection techniques are a mandatory preprocessing step before the classification process. Two hybrid feature selection approaches are proposed and applied on five breast cancer datasets, namely, Van’t veer, Chin, Chowdary, Gravier and West. First approach combined mutual information, Least Absolute Shrinkage and Selection Operator and genetic algorithm (MI-LASSO-GA). The second approach combined mutual information, Least Absolute Shrinkage and Selection Operator and particle swarm optimization (MI-LASSO-PSO). The following classifiers with 5-fold cross validation are used to assess the proposed approaches: Support Vector Machine (SVM), k-Nearest Neighbor (KNN), Random Forest (RF), Logistic Regression (LR), and extreme gradient boosting (XGBOOST). Statistical measures of accuracy, precision, recall and f1 metrics are used for performance analysis. MI-LASSO-PSO could produce subsets with fewer features than subsets produced by MI-LASSO-GA for all datasets except Chin. The average performance