Identifying Breast Cancer through the Application of Machine Learning Algorithms: A Comprehensive Exploration of Techniques and Methods for Detection and Diagnosis
Abstract
Cancer is a leading cause of mortality globally, with breast cancer posing a significant threat to women's health worldwide. Early detection is key to effective treatment. This study employs machine learning techniques, specifically Multilayer Perceptron (MLP) and Support Vector Machine (SVM) classifiers, to classify breast cancer data. Utilizing SVMRFE for dimensionality reduction, the study aims to identify the smallest subset of features for improved classification of benign and malignant tumors. By analyzing the Wisconsin Breast Cancer (WBC) dataset, this research seeks to optimize feature selection methods to enhance classification accuracy. Results indicate that MLP classifier achieves higher accuracy rates post feature selection. Comparative analysis of SVM and Artificial Neural Network underscores the efficacy of feature selection techniques in improving classification performance.
Keywords
Download Options
Introduction
The advancement of digital diagnostics has been driven by the imperative to support medical professionals in decision-making. Initially applied in healthcare for tasks like electrocardiograms, their usage has expanded to encompass ultrasounds and other traditional diagnostic methods. Traditionally, disease detection and monitoring heavily relied on healthcare professionals. However, the growing number of patients requiring continuous assessment has propelled technical advancements in automated systems. Converting qualitative data into quantitative measures is pivotal in addressing diagnostic challenges. Cancer, a broad term encompassing a multitude of diseases affecting various body parts, is characterized by the rapid proliferation of abnormal cells that surpass their normal boundaries, infiltrating adjacent tissues and potentially spreading to other organs—a process known as metastasis. Metastases are the primary cause of cancer-related deaths worldwide. Breast cancer, specifically, ranks as the second leading cause of cancer mortality among women aged 40 to 55, with an estimated 1.2 million new cases diagnosed annually according to projections by the World Health Organization.
Cancer is characterized by uncontrolled cell growth within the body, often named after the affected body structure. Breast cancer, notorious for its high mortality rate in women, manifests as rapidly dividing cells forming masses within the breast— referred to as tumors. Tumors are categorized as benign or malignant, with malignant tumors invading healthy tissues and potentially spreading to other organs, causing further damage. Breast cancer specifically denotes a malignant tumor within the breast.
Consequently, numerous studies have focused on early cancer detection, given its detrimental impact on human health. This study aims to diagnose cancer using the Wisconsin Diagnostic Breast Cancer (WDBC) dataset."[1][2][8][9].
Conclusion
The precise early detection of breast cancer cells can be facilitated through the utilization of AI techniques, potentially reducing healthcare costs and expediting treatment initiation for patients. This paper explores the application of Multilayer Perceptron (MLP) and Support Vector Machine (SVM) algorithms for diagnostic and predictive assessment of breast cancer. Consequently, feature selection methods have become imperative in several studies. In this research, a comparative analysis was conducted based on SVM-RFE-based feature selection algorithms to predict the risks associated with Breast Cancer Wisconsin (Diagnostic) infection. We proposed an SVM-RFE based feature selection method for classification tasks, aiming to integrate SVM-RFE computation with MLP and SVM algorithms to enhance classifier accuracy. Our experimental results indicate that reusing features eliminated during the SVM-RFE process can improve the performance of the MLP classifier. MLP is found to outperform SVM by providing higher prediction accuracy."