Breast cancer is a leading cause of mortality worldwide, emphasizing the need for accurate and timely prediction of risk factors. This study proposes a novel approach using deep learning algorithms to predict breast cancer. The self-adaptive sea lion optimization-based recurrent neural network (SA-SLnO-RNN) is developed to select optimal features that enhance prediction accuracy. Evaluated on the breast cancer Wisconsin diagnostic (BCWD) dataset, the SA-SLnO-RNN achieves superior performance compared to existing methods, with accuracy, precision, recall, specificity, and F1-score values of 98.82%, 98.78%, 98.71%, 99.06%, and 98.75%, respectively. The proposed model’s self-adaptive nature prevents overfitting, while the RNN with optimization algorithm enables quick convergence and captures complex temporal dependencies, ensuring early and accurate breast cancer prediction.