Stroke Prediction using Machine Learning

The main cause of major long-term disability worldwide is still ischemic stroke, which is a common neurological condition. For efficient intervention and therapy, the selection of factors related to stroke prognosis is extremely valuable. The International Stroke Trial (IST) dataset's prognostic characteristics for stroke were chosen for this study using an integrated machine learning technique. In this study, we looked at some typical issues with feature selection and prediction in datasets from the medical field. In the beginning, the Shapiro-Wilk algorithm was used to rank the significance of the features, and the Pearson correlations between the features were examined. Then, to choose robust features, we utilized Recursive Feature Elimination with Cross-Validation (RFECV), which included lincar SVC, Random-Forest-Classifier, Extra-Trees-Classifier, AdaBoost-Classifier, and Multinomial-Nave-Bayes- Classifier as estimators, in that order. Ischemic stroke, a common neurological disorder, continues to be the primary cause of significant long-term impairment globally. The choice of parameters associated to stroke prognosis is particularly valuable for effective management and therapy. The prognostic features of stroke from the International Stroke Trial (IST) dataset were selected for this study using an integrated machine learning approach. In this paper, we examined a few common problems with feature selection and prediction in medical datasets. The relevance of the features were first ranked using the Shapiro- Wilk algorithm, then the Pearson correlations between the features were looked at. Keywords: Block Chain, Food Items, Visibility and Traceability Systems, Food-Related Issues, Bit coin SHA- 256, Consensus.