Summary - scikit-learn

In this group you learned the scikit-learn estimator API.

Summary: scikit-learn

In this group:

The shared interface: fit, predict, transform, score; models swap with one line.
Estimator roles: predictor, transformer; hyperparameters in the constructor, learned values with _.
clone, get_params, set_params.
predict_proba, decision_function; reshape(1, -1) for a single example.
The default score; custom scorers with get_scorer and make_scorer.

Exercise Comparison on synthetic data

Generate data with make_moons(n_samples=300, noise=0.25, random_state=0) and split it with test_size=0.3, random_state=0. Train LogisticRegression() and SVC(); on each line print the class name and the test accuracy with round(..., 3).

Show example solution

from sklearn.datasets import make_moons
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC

X, y = make_moons(n_samples=300, noise=0.25, random_state=0)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)
for m in [LogisticRegression(), SVC()]:
    m.fit(X_train, y_train)
    print(type(m).__name__, round(m.score(X_test, y_test), 3))

Test yourself

Why does cross-validation use <code>clone</code> in every fold?
- To share the weights
- It doesn't
- So every fold starts with a clean, untrained model
- Because it is fast
Show answer

So every fold starts with a clean, untrained model — What was learned in the previous fold must not affect the next fold's score.

Get the F1 scorer by name.

olcut = ____("f1_macro")

Show answer

olcut = get_scorer("f1_macro")

Match the tools.

make_classification

Synthetic data
clone

Clean copy
decision_function

Raw score
make_scorer

Custom metric

Next lesson: What is a Pipeline?

We Are Enhancing Your Experience

Cookie Preferences

Strictly Necessary Cookies

Analytical Cookies