[Retracted] An Effective Machine Learning-Based Model for an Early Heart Disease Prediction

Algorithm 1

Pseudocode of the proposed MLbPM.

1: Load the HD dataset

2: Identify total number of patient records PR in

3: Find out nominal features NF in

4: fordo

5: fordo

6:     Apply one-hot encoding on each NF

7: end for

8:   Return

9: end for

10: Partition into two sets: training TR and testing TS data

11: Select base classifier algorithms CA

12: fordo

13:   Consider split ratio 80 : 20 SR1 and 70 : 30 SR2

14:   Predict the result of on TR with

15:   Predict the result of on TR with SR2

16:   Tune the hyperparameter of and go to step 12

17: end for

18: fordo

19:   Evaluate the performance of on TS

20:   Predict the result of on TS, save the results

21: end for

22: Get in step 8 and go to step 23

23: Identify numeric features NC in

24: fordo

25:   Identify scaling methods SM

26:   Scale each NC with a SM

27:   Return

28: end for

29: Execute steps 10 to 17 for , then go to step 30

30: fordo

31:   Evaluate the performance of each classifier on TS

32:   Predict the result of each classifier on TS, save the results

33: end for

34: Select and validate the best model