python-scikit-learnHow to define threshold for logistic regression

In order to define threshold for logistic regression we have to use predict_proba() method which predicts probabilities for each class/object and pick only classes with higher value than set by us:

from sklearn import datasets, linear_model, model_selection

X, y = datasets.load_iris(return_X_y=True)
X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y)

model = linear_model.LogisticRegression()
model.fit(X_train, y_train)
y_pred_prob = model.predict_proba(X_test)

threshold = 0.9
y_pred = []
for probs in y_pred_prob:
  cls = None
  for c, p in enumerate(probs):
    if p >= threshold:
      cls = c
      break
  y_pred.append(cls)ctrl + cgithub

`from sklearn import` import module from scikit-learn	`load_iris` loads Iris dataset
`model_selection.train_test_split` splits given `X` and `y` datasets to test (25% of values by default) and train (75% of values by default) subsets	`.LogisticRegression()` creates logistics regression model
`.fit(` train model with a given features and target variable dataset	`.predict_proba(` predict classes probabilities for a given sample
`y_pred_prob` list of predicted classes probabilities for each object	`threshold` minimum probability (confidence) level for predicting a class
`if p >= threshold:` pick only those classes which have higher probability than `threshold` value	`y_pred = []` will contain predicted classes with a given threshold value (and `None` if no confident class was found for an object)

Usage example

from sklearn import datasets, linear_model, model_selection

X, y = datasets.load_iris(return_X_y=True)
X_train, X_test, y_train, y_test = model_selection.train_test_split(X, y)

model = linear_model.LogisticRegression()
model.fit(X_train, y_train)
y_pred_prob = model.predict_proba(X_test)

threshold = 0.9
y_pred = []
for probs in y_pred_prob:
  cls = None
  for c, p in enumerate(probs):
    if p >= threshold:
      cls = c
      break
  y_pred.append(cls)

print(y_pred)

output

[None, 0, 0, None, None, 0, 1, 1, None, 0, None, None, None, None, None, None, 0, None, 1, 0, 2, 0, 0, None, 1, 1, 2, 0, 0, 0, 2, None, None, None, 2, None, None, None]

More of Python Scikit Learn

See more codes...

python-scikit-learnHow to define threshold for logistic regression

Usage example

output

Related

More of Python Scikit Learn