Page 63 - FULL REPORT 30012024

2.9.3 Stroke Disease Detection and Prediction Using Robust

Learning Approaches

The study focuses on the development of a machine learning model for
predicting strokes based on physiological variables. The researchers used the

stroke prediction dataset which is the same as the dataset that will be use from
this project, which had 5110 rows and 12 columns. The dataset was

imbalanced, with 249 rows indicating a stroke risk and 4861 rows indicating

no stroke risk. Figure 2.17 illustrates the proposed block diagram for this work.

Figure 2.17 Proposed block diagram for the project
(Source: Tazin et al., 2022)

The researchers employed four machine learning algorithms: Random Forest

(RF), Decision Tree, Voting Classifier, and Logistic Regression. They
performed data preprocessing, including handling null values, label encoding,

and balancing the dataset. The models were trained using an 80:20 ratio of
training to testing data.

46