SVC: nastavitev hiperparametrov

1 Priprava podatkov

Izbrane spremenljivke (p < 0.10):

1. HospitalizationBeforeSurgeryMo
2. Age
3. BMI
4. CardiogenicSchockYN
5. Diabetes
6. DiabetesPerOsTherapie
7. DiabetesOnInsuline
8. PerifernoArterijskoObolenje
9. ExtracardiacArteriopathy
10. PsychoSyndrome
11. TherapyRelevantPsychoSyndrome
12. PreoperativeInfectionYN
13. KongestiveHeartFailure
14. EjectionFractionEF
15. EF50
16. AtrialFibrillationYN
17. ChronicLungDiseaseYN
18. CockcraftGaultIndexPreop
19. ACEInhibitors
20. IABPPreoperatively
21. DurationOfTheOperation
22. NumberOfGrafts
23. PericardDrainage
24. RethoracotomyYN
25. CoagulationDisorder
26. Cardioversion
27. SumOtherInfectYN
28. AcuteKidneyFailure
29. TotalDrainage
30. NumberOfPlasmaUnits
31. Transfusion
32. MoreThan2UnitsOfErythrocytes
33. RespiratoryFailureYN
34. ProlongedMechanicalVentilation
35. Reintubation
36. NumberOfReintubations
37. Tracheotomy
38. AorticClampingTime
39. BypassOperationTime
40. AorticCalcificatio
41. LeukocytesFirstPostoperativeDa
42. LeukocytesSecondPostoperativeD
43. HbPreop12GDl
44. HbPreoperativelyGDl
45. PoorGlycemicControlPrediabetes
46. ITA
47. BIMA
48. SaEtAlCreatinine226MgDdlOrPost
49. TiesselFibrinGlueMl
50. GFRLaurisProdop60NotNormal
51. GFRLaurisPostop1stDay60NotNorm
52. GFRLaurisPostop2ndDay60NotNorm
53. PleuralEffusion

Target: DSWI01 | Predictors: 53 | Samples: 4023

Izvedeno skaliranje s standardizacijo numeričnih spremenljivk in “one-hot encoding” za kategorijske spremenljivke.

2 Nastavitev hiperparametrov

Najboljša nastavitev:

• model__C: 3.0
• model__class_weight: None
• model__gamma: scale
• model__kernel: rbf

Best roc_auc: 0.918

Top 10 kombinacij (CV):

params_str	mean_test_accuracy	mean_test_f1	mean_test_roc_auc
model__C=3.0, model__class_weight=None, model__gamma=scale, model__kernel=rbf	0.942	0.653	0.918
model__C=3.0, model__class_weight=None, model__gamma=auto, model__kernel=rbf	0.942	0.653	0.918
model__C=3.0, model__class_weight=balanced, model__gamma=scale, model__kernel=rbf	0.932	0.687	0.909
model__C=3.0, model__class_weight=balanced, model__gamma=auto, model__kernel=rbf	0.932	0.686	0.909
model__C=1.0, model__class_weight=None, model__gamma=scale, model__kernel=rbf	0.932	0.544	0.907
model__C=1.0, model__class_weight=None, model__gamma=auto, model__kernel=rbf	0.932	0.542	0.907
model__C=1.0, model__class_weight=balanced, model__gamma=scale, model__kernel=rbf	0.912	0.631	0.906
model__C=1.0, model__class_weight=balanced, model__gamma=auto, model__kernel=rbf	0.913	0.632	0.906
model__C=0.5, model__class_weight=balanced, model__gamma=scale, model__kernel=rbf	0.900	0.599	0.903
model__C=0.5, model__class_weight=balanced, model__gamma=auto, model__kernel=rbf	0.900	0.599	0.903

CV setup: StratifiedKFold with 5 folds (shuffle=True, random_state=1974).