Summary - Regression

In this group you learned linear regression, model training, and prediction.

Summary: regression

In this group:

Regression predicts a continuous number; models are named ...Regressor.
Linear regression fits the line y = w·x + b by least squares; coef_ and intercept_ are interpretable.
fit learns the parameters, predict predicts with two-dimensional input.
It is measured with MAE, MSE, RMSE and R²; measuring on test data is essential.

m = LinearRegression().fit(X_train, y_train)
p = m.predict(X_test)
print(mean_absolute_error(y_test, p), r2_score(y_test, p))

Exercise End-to-end regression

Split the data with test_size=0.25, random_state=0 and train a LinearRegression. On the test set, print the MAE with round(..., 2) and the R² with round(..., 3) on separate lines.

Show example solution

import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_absolute_error, r2_score

rng = np.random.default_rng(0)
X = rng.uniform(0, 10, size=(40, 1))
y = 4 * X[:, 0] + 3 + rng.normal(0, 1, size=40)

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25, random_state=0)
m = LinearRegression().fit(X_train, y_train)
p = m.predict(X_test)
print(round(mean_absolute_error(y_test, p), 2))
print(round(r2_score(y_test, p), 3))

Test yourself

What can be said of a model with an R² of -0.2?
- It is very good
- The error is 20%
- R² can't be negative
- It is worse than always predicting the mean
Show answer

It is worse than always predicting the mean — A negative R² shows the model makes more error than the simple mean prediction.
Get R² on the test set directly from the model.
```
r2 = m.____(X_test, y_test)
```
Show answer
```
r2 = m.score(X_test, y_test)
```
Match the tools.

LinearRegression

Linear model
coef_

Weights
mean_squared_error

MSE
r2_score

R²

Next lesson: What is KNN?

We Are Enhancing Your Experience

Cookie Preferences

Strictly Necessary Cookies

Analytical Cookies