Program Type

Graduate

Faculty Advisor

Dr. Jerry Wood and Dr. Tolga Ensari

Document Type

Poster

Location

Face-to-face

Start Date

18-4-2024 4:00 PM

Abstract

Breast cancer is one of the foremost causes of death amongst women worldwide. Breast tumours are characteristically classified as either benign (non-cancerous) or malignant (cancerous). Benign tumours do not spread external side of the breast and are not fatal, whereas malignant tumours can metastasize and be incurable if untreated. Rapidly and accurate diagnosis of malignant tumours is significant for efficient treatment and advanced outcomes. In 2022, breast cancer claimed 670 000 lives worldwide. Women without any particular risk factors other than age and sex account for half of all cases of breast cancer. In 157 out of 185 nations, breast cancer was the most frequent cancer among women in 2022. Worldwide, breast cancer affects people in every nation. Men are affected by breast cancer at a rate of 0.5–1% [1]. One prevalent and vigorous machine learning algorithm that has been significantly used for classification undertakings is the Support Vector Machine (SVM). SVM is a supervised learning model that forms an optimal hyperplane or decision boundary to maximize the margin concerning classes. It can handle high-dimensional and non-linear data by using kernel functions to map the data into a higher-dimensional space where it develops linearly separable. By exercising the SVM algorithm, I aim to leveraging its adeptness to efficiently handle high-dimensional data and portray complex, non-linear relationships involving features and class labels. The SVM's flexibility in using different kernel functions acknowledges for modelling diverse decision boundaries, theoretically steering to advanced classification accuracy.

In this study, I will investigate the performance of the SVM classifier on the Wisconsin Breast Cancer dataset and compare it with the formerly reported results using KNN and NB classifiers. Furthermore, I will probe the effect of different kernel functions and hyperparameter tuning on the SVM's performance to enhance its classification capabilities.

The results of this research will provide perceptions into the effectiveness of the SVM algorithm for breast cancer diagnosis and influence on the development of accurate and trustworthy machine learning-based diagnostic tools. Precise classification of benign and malignant tumours can assist healthcare professionals in determining the applicable course of cure and managing for patients, ultimately recovering clinical outcomes. Healthcare professionals in determining the appropriate course of treatment and management for patients.

Share

COinS
 
Apr 18th, 4:00 PM

Breast Cancer Classification with Machine Learning

Face-to-face

Breast cancer is one of the foremost causes of death amongst women worldwide. Breast tumours are characteristically classified as either benign (non-cancerous) or malignant (cancerous). Benign tumours do not spread external side of the breast and are not fatal, whereas malignant tumours can metastasize and be incurable if untreated. Rapidly and accurate diagnosis of malignant tumours is significant for efficient treatment and advanced outcomes. In 2022, breast cancer claimed 670 000 lives worldwide. Women without any particular risk factors other than age and sex account for half of all cases of breast cancer. In 157 out of 185 nations, breast cancer was the most frequent cancer among women in 2022. Worldwide, breast cancer affects people in every nation. Men are affected by breast cancer at a rate of 0.5–1% [1]. One prevalent and vigorous machine learning algorithm that has been significantly used for classification undertakings is the Support Vector Machine (SVM). SVM is a supervised learning model that forms an optimal hyperplane or decision boundary to maximize the margin concerning classes. It can handle high-dimensional and non-linear data by using kernel functions to map the data into a higher-dimensional space where it develops linearly separable. By exercising the SVM algorithm, I aim to leveraging its adeptness to efficiently handle high-dimensional data and portray complex, non-linear relationships involving features and class labels. The SVM's flexibility in using different kernel functions acknowledges for modelling diverse decision boundaries, theoretically steering to advanced classification accuracy.

In this study, I will investigate the performance of the SVM classifier on the Wisconsin Breast Cancer dataset and compare it with the formerly reported results using KNN and NB classifiers. Furthermore, I will probe the effect of different kernel functions and hyperparameter tuning on the SVM's performance to enhance its classification capabilities.

The results of this research will provide perceptions into the effectiveness of the SVM algorithm for breast cancer diagnosis and influence on the development of accurate and trustworthy machine learning-based diagnostic tools. Precise classification of benign and malignant tumours can assist healthcare professionals in determining the applicable course of cure and managing for patients, ultimately recovering clinical outcomes. Healthcare professionals in determining the appropriate course of treatment and management for patients.