<p>Breast cancer is a prevalent form of cancer among women, and its detection is often a complex and challenging process due to a multitude of factors. This study introduces an innovative approach to breast cancer prediction using 1D convolutional neural networks (1D-CNN). Our study revolves around the development and optimization of a 1D-CNN model, leveraging the UCI Wisconsin Diagnostic Breast Cancer Data (WDBC). We begin our exploration with a baseline 1D-CNN model, establishing the foundation for our research. Subsequently, we enhance the model’s performance through weight optimization using a genetic algorithm, resulting in the BC-GAWOCNN. The third experiment delves into hyperparameter optimization with the BC-GAHOCNN model, pushing the model’s capabilities to achieve outstanding results in breast cancer detection. Notably, our optimized model BC-GAHOCNN, demonstrates a remarkable classification accuracy of 99%, precision, recall, and F1-score, surpassing performance compared to other existing models. Additionally, the area under the curve was 0.99, highlighting the model’s strong ability to distinguish between positive and negative cases. This high level of performance was obtained by using the mean of three cross-validation F1-scores as the fitness function of GA. Moreover, we obtained excellent classification results when employing mean accuracy as the fitness function, with a precision of 99%, an accuracy of 98%, recall and F1-scores of 98%. Our research aims to bring new hope and efficiency to both patients and healthcare professionals in the fight against this devastating disease.</p>