Accurate credit rating by banks is crucial for enhancing long-term profitability through effective customer acquisition and management of existing credit limits, as well as for contributing to the economy through systematic and efficient capital allocation. Furthermore, identifying suitable credit opportunities that benefit customers, banks, and all stakeholders enhances reliability. Additionally, it is desired that credit rating prediction models employed by banks be explainable to financial authorities. Although machine learning-based systems demonstrate improved performance in various tasks, some advanced models may suffer from deficiencies in transparency, reliability, and explainability. Therefore, the application of Explainable AI (XAI) methods in banking machine learning techniques enhances the interpretability, comprehensibility, and reliability of prediction models. This study aims to identify the features influencing decisions within the banking credit system and to compare the results obtained from artificial intelligence by ranking the most significant features in credit rating. A prediction model was developed by applying Particle Swarm Optimization (PSO) to an Artificial Neural Network (ANN), and the machine learning technique that best explains the model was determined. The explainability of the prediction model was assessed by comparing the results of Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP). These methods were evaluated based on their ability to explain the local interpretability of the model and the contribution of each feature to the prediction outcome. Additionally, the findings related to credit limit prediction were presented along with a discussion of the strengths and weaknesses of interpretability. Demonstrating transparently how machine learning models arrive at their recommendations for credit limits not only enhances customer satisfaction but also ensures that financial institutions comply with legal requirements and fosters trust among users.