<![CDATA[Breast cancer is one of the diseases with a high prevalence and is a leading cause of death among women. Early detection is crucial in improving patient survival rates. However, a major challenge in diagnosis using machine learning methods is the high dimensionality of the data, which can lead to overfitting and reduced interpretability of the model. This study proposes a new approach to improve breast cancer prediction accuracy by using a combination of Genetic Algorithm + Neural Network (GA + NN). The dataset used is the Breast Cancer Wisconsin (Diagnostic) Data Set, consisting of 569 samples with 32 numerical features that describe the characteristics of tumor cells. The experimental results show that the GA + MLP method achieved the highest accuracy of 99.42%, outperforming the benchmark model using PCA and logistic regression with an accuracy of 97.37%. This approach demonstrates that GA-based feature selection can improve prediction accuracy while reducing model complexity, making it more efficient for medical applications.]]>
Copyrights © 2025
