from sklearn.svm import SVC
from sklearn.svm import LinearSVC
import numpy as np
[docs]class NaivelyCalibratedLinearSVC(LinearSVC):
"""LinearSVC with `predict_proba` method that naively scales
`decision_function` output."""
name = "lscv"
[docs] def fit(self, X, y):
super().fit(X, y)
df = self.decision_function(X)
self.df_min_ = df.min()
self.df_max_ = df.max()
[docs] def predict_proba(self, X):
"""Min-max scale output of `decision_function` to [0,1]."""
df = self.decision_function(X)
calibrated_df = (df - self.df_min_) / (self.df_max_ - self.df_min_)
proba_pos_class = np.clip(calibrated_df, 0, 1)
proba_neg_class = 1 - proba_pos_class
proba = np.c_[proba_neg_class, proba_pos_class]
return proba
[docs]class NaivelyCalibratedSVC(SVC):
"""LinearSVC with `predict_proba` method that naively scales
`decision_function` output."""
name = "svc"
hyperparameters = [{"kernel": ["rbf"], "gamma": ["scale", "auto"]}]
[docs] def fit(self, X, y):
super().fit(X, y)
df = self.decision_function(X)
self.df_min_ = df.min()
self.df_max_ = df.max()
[docs] def predict_proba(self, X):
"""Min-max scale output of `decision_function` to [0,1]."""
df = self.decision_function(X)
calibrated_df = (df - self.df_min_) / (self.df_max_ - self.df_min_)
proba_pos_class = np.clip(calibrated_df, 0, 1)
proba_neg_class = 1 - proba_pos_class
proba = np.c_[proba_neg_class, proba_pos_class]
return proba
NaivelyCalibratedSVCLinear = NaivelyCalibratedLinearSVC(C=1.0)
NaivelyCalibratedSVCRBF = NaivelyCalibratedSVC(kernel="rbf", gamma=0.7, C=1.0)
NaivelyCalibratedSVCRBF.name = "rbf_svc"
NaivelyCalibratedSVCPolynomial = NaivelyCalibratedSVC(
kernel="poly", degree=5, C=1.0)
NaivelyCalibratedSVCPolynomial.name = "poly_svc"