Stroke prediction using 1DCNN with ANOVA

Stroke and heart disease are among the most com-mon outcomes of hypertension. Each year, disease, stroke, other cardiovascular disorders claim lives more than 877,500 people in United States, making them first fifth leading causes death, so being able to pre- dict early helps save lives. A lot research has been done reach this goal. Machine learning models mostly used for purpose. For time study, we have Deep Learning (DL) model, i.e., one dimen- sional convolutional neural network (1D CNN) . In extracted important features using Analysis variance (ANOVA) method. Then data set with new that came up was given model. compare all machine algorithms—K-Nearest Neighbors (KNN), Sup- port Vector (SVM), Logistic Regression (LR), Random Forest Classi- fier (RF), Gradient Boosting Clas-sifier (XGB), LoLight gradient boosting classifier (LGBM)—with 1DCNN. Recall, F1 score, accuracy, precision some confusion metrics assess effectiveness results.The results show when on reprocessed data, proposed model performs best is 98% accurate.