Algorithm 2.

Showcasing the Major Working Steps of RDKST.

1: Inputs: Dataset, $D={\sum }_{i=1}^M\left(X_i,Y_i\right)$, Meta classifiers = LG

2: Outputs: Classify whether the thyroid is affected or not

3: $D^{\left(a\right)}$ ← D.drop ([TBG], axis = columns)

4: $D^{\left(b\right)}$ ← MedianImpute {$D^{\left(a\right)}$}

5: $D^{\left(c\right)}$ ← LabelEncoder {$D^{\left(b\right)}$}

6: $D^{\left(d\right)}$ ← SMOTEENN {$D^{\left(c\right)}$}

7: $X_i,Y_i$ ← input{$D^{\left(d\right)}$(N × M matrix)}, output (N × 1 vector)

8: ${Xtr}_7,{Ytr}_7,{Xte}_3,{Yte}_3$ ← TrainTestSplit ($X_i,Y_i,0.3$)

9: ${Xtr}_8,{Ytr}_8,{Xte}_2,{Yte}_2$ ← TrainTestSplit ($X_i,Y_i,0.2$)

10: ${Xtr}_9,{Ytr}_9,{Xte}_1,{Yte}_1$ ← TrainTestSplit ($X_i,Y_i,0.1$)

11: while (execute − different − TrainTestSets) do

12: ${BC}^{\left(1\right)}$ ← RandomForest (${Xtr}_i,{Ytr}_i,{Yte}_i$)

13: ${BC}^{\left(2\right)}$ ← DecisionTree (${Xtr}_i,{Ytr}_i,{Yte}_i$)

14: ${BC}^{\left(3\right)}$ ← KNearestNeighbors (${Xtr}_i,{Ytr}_i,{Yte}_i$)

15: end while

16: ${Base}_{pred}$ ← concatenate (${BC}^1,{BC}^2,and{BC}^3$)

17: for$i=1;i<M;i++$do

18: Apply ${Base}_{pred}$ to classify training instances $X_i$

19: $X_i$ ← ${Base}_{pred}$ ($X_i$)

20: $Inpu{t}_{\mathit{LG}}$ ← ($X_i^{\delta }$, $Y_i$), where $X_i^{\delta }$ ← ($X_{1i},\dots .X_{Mi}$)

21: end for

22: RDKST ← LG {$Inpu{t}_{\mathit{LG}}$}

23: Result ← RDKST. predict (New−sample)

24: ReturnResult