Algorithm 2: The algorithm for the Cleveland dataset used in this study.

Input: symptoms

Output: predict heart disease present or not present

1. If (the model has not been trained), then

2. Dataset load;

3. Correlation of data;

4. Check outliers;

5. Remove outliers;

6. Split x and y;

7. Train (80%), test (20%);

8. Load pre-trained model;

9. Educate the model;

10. Save the model that has been trained.

11. Loads trained model if everything else fails;

12. Validate the model using the test data set;

13. Confusion metrics and plot graphs.