This study presents an interpretable approach for heart stroke prediction using explainable AI techniques. The research aims to develop a reliable and understandable model for healthcare professionals to use in clinical decision-making. The proposed model incorporates resampling, data leakage prevention, and feature selection to enhance model accuracy and interpretability. The Stroke Prediction Dataset, containing 11 attributes, was used to train and evaluate the model. The model achieved an accuracy of 95%, demonstrating its effectiveness in predicting heart strokes. The study addresses the challenge of class imbalance in the dataset and employs SMOTE Tomek to generate a balanced dataset. Data leakage was also addressed by ensuring that the training data was split before any transformations. Feature selection was performed using mutual information and chi-square tests to identify the most relevant features for prediction. The model was evaluated using various performance metrics, including accuracy, precision, recall, specificity, and F1-score. The results showed that resampling techniques improved recall and F1-scores, although they slightly reduced accuracy and precision. The study also emphasizes the importance of model interpretability, using techniques such as LIME and permutation importance to explain the model's predictions. These techniques provide insights into the factors influencing specific predictions, making the model more understandable for healthcare professionals. The research contributes to the field of healthcare by offering a reliable and interpretable tool for heart stroke prediction, which can help in early diagnosis and prevention of stroke-related conditions. The study highlights the potential of explainable AI in improving the accuracy and interpretability of predictive models in clinical settings.This study presents an interpretable approach for heart stroke prediction using explainable AI techniques. The research aims to develop a reliable and understandable model for healthcare professionals to use in clinical decision-making. The proposed model incorporates resampling, data leakage prevention, and feature selection to enhance model accuracy and interpretability. The Stroke Prediction Dataset, containing 11 attributes, was used to train and evaluate the model. The model achieved an accuracy of 95%, demonstrating its effectiveness in predicting heart strokes. The study addresses the challenge of class imbalance in the dataset and employs SMOTE Tomek to generate a balanced dataset. Data leakage was also addressed by ensuring that the training data was split before any transformations. Feature selection was performed using mutual information and chi-square tests to identify the most relevant features for prediction. The model was evaluated using various performance metrics, including accuracy, precision, recall, specificity, and F1-score. The results showed that resampling techniques improved recall and F1-scores, although they slightly reduced accuracy and precision. The study also emphasizes the importance of model interpretability, using techniques such as LIME and permutation importance to explain the model's predictions. These techniques provide insights into the factors influencing specific predictions, making the model more understandable for healthcare professionals. The research contributes to the field of healthcare by offering a reliable and interpretable tool for heart stroke prediction, which can help in early diagnosis and prevention of stroke-related conditions. The study highlights the potential of explainable AI in improving the accuracy and interpretability of predictive models in clinical settings.