Gaussian Naive Bayes for Brain Tumor Classification

This study presents a robust classification framework for brain tumor diagnosis using Gaussian Naive Bayes. The method was selected for its simplicity, efficiency with high-dimensional data, and suitability for continuous features, distinguishing it from complex models like neural networks and decision trees. Unlike other approaches requiring intensive computation and hyperparameter tuning, Gaussian Naive Bayes provides a balance of interpretability and performance. This method effectively handles missing data through imputation and standardization, ensuring reliable analysis. The model was evaluated using cross-validation, achieving a consistent accuracy of approximately 65 %, with performance assessed across precision, recall, and F1 scores for various tumor types. The confusion matrix highlights the strengths and areas for improvement in classification accuracy. This approach demonstrates significant potential for clinical applications, offering a quick and accessible diagnostic tool for healthcare professionals.