Revisemos el comportamiento de este kernel para varios valores del parámetro ɣ:
gammas = [0.001, 0.01, 0.1, 1]
for g in gammas:
print("Gamma:", g)
model = SVC(kernel = "sigmoid", gamma = g, coef0 = 0)
model.fit(X.values, y)
show_boundaries(model, X.values, None, y, None, labels = iris.species.unique())
for g in gammas:
print("Gamma:", g)
model = SVC(kernel = "sigmoid", gamma = g, coef0 = 0)
model.fit(X.values, y)
show_boundaries(model, X.values, None, y, None, labels = iris.species.unique())
Gamma: 0.001
Gamma: 0.01
Gamma: 0.1
Gamma: 1
