Research Article
A Constructive Data Classification Version of the Particle Swarm Optimization Algorithm
Pseudocode 1
Particle Swarm Clustering (PSC).
1. procedure X, V, P, g, , PLABELS = PSC (Y, max, , , CLABELS) | 2. Y//dataset | 3. initialize X//initialize at random each particle 0,1 | 4. initialize //initialize at random each −max, max | 5. initialize dist | 6. | 7. while stopping criterion is not met | 8. for to //for each input datum | 9. for to //for each particle | 10. dist () = distance () | 11. end for | 12. = index (min (dist)) | 13. PLABELS = label (, CLABELS ())//predicted label | 14. if distance () < distance () | 15. | 16. end if | 17. if distance () < distance () | 18. | 19. end if | 20. = + 1 () + 2 () + 3 () | 21. −max, max | 22. () = () + () | 23. 0, 1 | 24. end for | 25. for to | 26. if ( = win )//particles did not | 27. = ω * () + φ 4 (x most_win − ()) | 28. ∈ −max, max | 29. () = () + | 30. end if | 31. end for | 32. | 33. | 34. Test the stopping criterion | 35. end while | 36. end procedure |
|