Algorithm 2 WC Feature Selection method

1. Set Mpop, Msr, dimax, Max_Iteration. 2. Execute Equations (11) and (12)//to determine the number of streams flowing into rivers and sea 3. k = Features (WSN-DS) 4. Mpop (1:k) = random initial population 5. Set the classifier algorithm (DT, KNN, SVM, DL, NB) 6. Calculate the fitness function for Mpop (1:k) using Equation (10) 7. Sort fitness values in descending order 8. F_sea = best fitness for initial populations 9. F_river = best next fitness after the F_sea 10. F_stream = Execute Equation (13)//to determine the number of streams flow to their corresponding rivers and sea, which is considered the best fitness after F_river 11. t = 0 12. while (t < Max_Iterration) do 13. for i = 1: Mpop do 14.   new_stream = Execute Equations (15) and (16)//to find new stream flows 15.    F_new_stream = Execute Equation (10) for new_stream 16.   if F_new_stream < F_river then 17.    River = new_stream 18.   if F_new_stream < F_sea then 19.    Sea = new_stream 20.   end if//line 16 21.   end if//line 18 22.   new_river = Execute Equation (17)//to find new river flows 23.    F_new_river = Execute Equation (10) for new_river 24.   if F_new_river < F_sea then 25.    Sea = new_river 26.    end if//line 24 27.   end for//line 13 28.  for i = 1: Msr do 29.   if (distance (Sea and River) < dimax) or (rand < 0.1) then 30.   New_stream = Execute Equation (18) 31.   end if//line 29 32.  end for//line 28 33. t = t + 1 34. end while 35. print the performance accuracy and number of selected features