Breast Cancer detection using Jax Based machine learning models

Kadam, Nikhil (2023) Breast Cancer detection using Jax Based machine learning models. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (3MB) | Preview

Abstract

This study investigates the application of JAX-based machine learning models for early breast cancer detection, utilizing the Breast Cancer Wisconsin (Diagnostic) dataset and a Breast Ultrasound Image dataset. The research evaluates a range of models including JAX-based Logistic Regression and Neural Networks, alongside traditional Logistic Regression and Neural Network using Scikit Learn, and deep learning architectures such as CNN, InceptionNet, DenseNet121, and VGG19.

Key findings reveal that the Scikit Learn-based Logistic Regression model excels in the structured Wisconsin dataset, achieving approximately 99% accuracy, while the Neural Network model also shows high efficiency. JAX-based models, however, exhibit mixed results, with the Neural Network particularly underperforming, indicating challenges in complex model implementations within JAX. In the image dataset evaluation, VGG19 outperforms others with about 77% accuracy, highlighting its strength in image-based data analysis. In contrast, CNN and DenseNet121 show moderate effectiveness, and InceptionNet falls behind, suggesting a potential mismatch with the dataset’s features.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Anant, Aaloka UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > Life sciences > Medical sciences > Pathology > Tumors > Cancer Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	09 May 2025 13:11
Last Modified:	09 May 2025 13:11
URI:	https://norma.ncirl.ie/id/eprint/7539

Actions (login required)

View Item