Feature Selection for High-Dimensional and Imbalanced Biomedical Data Based on Robust Correlation Based Redundancy and Binary Grasshopper Optimization Algorithm

. 2020 Jun 27;11(7):717. doi: 10.3390/genes11070717

Algorithm 2

1: For each feature in the training set D{

f_{1}

f_{2}

,...,

f_{n}

} and class

C

2: Initialize parameters,

S

\emptyset

R

\emptyset

k :

number of selected filters

k :

selected filters

W \in [R e l i e f F, F i s h e r s c o r e, C h i - S q u a r e]

Ensemble of filters

5: for

i

\in

{1, . . ., k} d o

6: for

j \in {1, \dots ., m} d o

Employ

W_{i}

to calculate the statistical scores of each gene

G_{i}

End of for

7: Rank genes based on its scores

G_{i}

in descending order and get a new gene list

W_{i} G_{j}

j = 1, \dots ., m

End of for

8: produce a new ranking list

R

by combining

k

filter methods

W_{i} G_{i}

i = 1, \dots ., k

using union operator, in order to consolidate the overlapping genes.