2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions
| Format: | General Document |
|---|
| _version_ | 1860798287864397824 |
|---|---|
| building | INTELEK Repository |
| collection | Online Access |
| collectionurl | https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 |
| copyright | Copyright©PWB2025 |
| country | Malaysia |
| date | 2024-11-05 07:07 |
| format | General Document |
| id | 16819 |
| institution | UniSZA |
| originalfilename | 16819_ce8aec496de600b.pdf |
| person | Muhamad Amierusyahmi Bin Zuhairi |
| recordtype | oai_dc |
| resourceurl | https://intelek.unisza.edu.my/intelek/pages/view.php?ref=16819 |
| sourcemedia | Server storage Scanned document |
| spelling | 16819 https://intelek.unisza.edu.my/intelek/pages/view.php?ref=16819 https://intelek.unisza.edu.my/intelek/pages/search.php?search=!collection3 General Document Malaysia Library Staff (Top Management) Library Staff (Management) Library Staff (Support) Terengganu Faculty of Informatics & Computing English application/pdf 1.6 www.ilovepdf.com Server storage Scanned document Universiti Sultan Zainal Abidin UniSZA Private Access Universiti Sultan Zainal Abidin Copyright©PWB2025 Dissertations, Academic Deep learning (Machine learning) 119 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions Muhamad Amierusyahmi Bin Zuhairi Optimizing Deep Learning Models This thesis explores the development and implementation of an enhanced deep learning model aimed at improving prediction accuracy for medical datasets, specifically focusing on breast cancer and diabetes. Leveraging two key numeric datasets—the Wisconsin Breast Cancer Diagnostic (WBCD) and a diabetes dataset—this research addresses several pressing challenges in medical diagnostics. These challenges include the critical need for high prediction accuracy, the mitigation of data imbalance issues, and the optimization of deep learning models for large, complex datasets. A novel three-layer deep learning architecture was developed for this purpose, incorporating multiple hidden layers and dropout regularization techniques to minimize overfitting and improve generalization. Data pre-processing played a significant role in this research, including techniques like data augmentation and normalization to enhance model performance. Specifically, the augmentation of malignant samples in the WBCD dataset addressed class imbalance, and Min-Max scaling was applied to both datasets to ensure uniformity of feature values. The model’s architecture was further optimized by adjusting parameters like dropout rates, batch sizes, and activation functions to achieve optimal performance. The results demonstrated substantial improvements in prediction accuracy, with the proposed model achieving 95-99% accuracy across both datasets, a significant enhancement over previously reported results. For the WBCD dataset, the deep learning architecture effectively captured key features related to malignancy, while for the diabetes dataset, the model demonstrated strong predictive power in identifying key risk factors for the disease. These findings underscore the potential of deep learning techniques to enhance diagnostic precision in medical applications, offering valuable insights for healthcare professionals in early disease detection and personalized treatment planning. This research contributes to the ongoing advancement of machine learning applications in healthcare, displaying the ability of deep learning models to not only improve prediction accuracy but also to address challenges related to data complexity, class imbalance, and overfitting. 2024-11-05 07:07 uuid:b69811f0-95e3-4377-afb1-da6fe80f2526 16819_ce8aec496de600b.pdf Breast Cancer and Diabetes Predictions Breast — Cancer — Diagnosis — Data processing Diabetes — Diagnosis — Data processing Medical informatics — Data processing Breast Cancer Prediction Diabetes Prediction Wisconsin Breast Cancer Diagnostic Dataset (WBCD) Artificial Intelligence in Medicine Healthcare Analytics Predictive Modeling Thesis |
| spellingShingle | 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions |
| state | Terengganu |
| subject | Dissertations, Academic Deep learning (Machine learning) Breast — Cancer — Diagnosis — Data processing Diabetes — Diagnosis — Data processing Medical informatics — Data processing |
| summary | This thesis explores the development and implementation of an enhanced deep learning model aimed at improving prediction accuracy for medical datasets, specifically focusing on breast cancer and diabetes. Leveraging two key numeric datasets—the Wisconsin Breast Cancer Diagnostic (WBCD) and a diabetes dataset—this research addresses several pressing challenges in medical diagnostics. These challenges include the critical need for high prediction accuracy, the mitigation of data imbalance issues, and the optimization of deep learning models for large, complex datasets. A novel three-layer deep learning architecture was developed for this purpose, incorporating multiple hidden layers and dropout regularization techniques to minimize overfitting and improve generalization. Data pre-processing played a significant role in this research, including techniques like data augmentation and normalization to enhance model performance. Specifically, the augmentation of malignant samples in the WBCD dataset addressed class imbalance, and Min-Max scaling was applied to both datasets to ensure uniformity of feature values. The model’s architecture was further optimized by adjusting parameters like dropout rates, batch sizes, and activation functions to achieve optimal performance. The results demonstrated substantial improvements in prediction accuracy, with the proposed model achieving 95-99% accuracy across both datasets, a significant enhancement over previously reported results. For the WBCD dataset, the deep learning architecture effectively captured key features related to malignancy, while for the diabetes dataset, the model demonstrated strong predictive power in identifying key risk factors for the disease. These findings underscore the potential of deep learning techniques to enhance diagnostic precision in medical applications, offering valuable insights for healthcare professionals in early disease detection and personalized treatment planning. This research contributes to the ongoing advancement of machine learning applications in healthcare, displaying the ability of deep learning models to not only improve prediction accuracy but also to address challenges related to data complexity, class imbalance, and overfitting. |
| title | 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions |
| title_full | 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions |
| title_fullStr | 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions |
| title_full_unstemmed | 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions |
| title_short | 2024_Optimizing Deep Learning Models For Accurate Breast Cancer And Diabetes Predictions |
| title_sort | 2024_optimizing deep learning models for accurate breast cancer and diabetes predictions |