COMPARATIVE ANALYSIS OF DIMENSIONALITY REDUCTION FOR BREAST CANCER USING MACHINE LEARNING AND DEEP LEARNING

Breast cancer is one of the leading causes of death among women worldwide. Accurate early detection is essential to improve patient survival rates. Therefore, an efficient and optimal detection method is needed. This study presents a comparative analysis between machine learning and deep learning models integrated with various dimensionality reduction techniques to improve the accuracy of breast cancer classification. The dimensionality reduction methods evaluated include Principal Component Analysis (PCA), t-Distributed Stochastic Neighbor Embedding (t-SNE), Independent Component Analysis (ICA), and Linear Discriminant Analysis (LDA). This study uses a dataset from the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC), which includes genetic and clinical data of breast cancer patients. Several classification algorithms are used in the evaluation, including Logistic Regression, Support Vector Machines (SVM), and Convolutional Neural Networks (CNN). Model performance is analyzed based on accuracy, precision, recall, and F1-score metrics. The results show that the LDA technique consistently produces better classification performance compared to other dimensionality reduction methods on various Machine Learning and Deep Learning models. The importance of choosing the right dimensionality reduction method in increasing the effectiveness of learning algorithms and more optimal, especially in the context of complex and high-dimensional medical data. The implications of this study can be used to develop a smarter decision support system in breast cancer diagnosis.