Polynomial regression pipeline in ML Python - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is polynomial regression?

Polynomial regression is a type of regression analysis where the relationship between the input variable and the output variable is modeled as an nth degree polynomial. It helps capture curved patterns in data.

beginner

Why do we use a pipeline in polynomial regression?

A pipeline helps combine multiple steps like transforming features into polynomial features and fitting a regression model into one sequence. This makes the process cleaner, easier to manage, and reduces errors.

intermediate

What does the PolynomialFeatures transformer do in a pipeline?

PolynomialFeatures creates new features by raising the original features to different powers up to the specified degree. This allows the model to learn nonlinear relationships.

beginner

How do you evaluate the performance of a polynomial regression model?

You can evaluate it using metrics like Mean Squared Error (MSE) or R-squared (R²). MSE measures average squared errors, while R² shows how well the model explains the data variance.

intermediate

What is the risk of using a very high degree polynomial in regression?

Using a very high degree polynomial can cause overfitting, where the model fits the training data too closely and performs poorly on new data. It captures noise instead of the true pattern.

What is the main purpose of PolynomialFeatures in a regression pipeline?

A To normalize the data

B To create new features by raising inputs to powers

C To reduce the number of features

D To split data into training and testing sets

PolynomialFeatures generates new features by raising the original features to different powers, enabling the model to learn nonlinear relationships.

Which metric is commonly used to measure the error of a polynomial regression model?

A Mean Squared Error (MSE)

B Accuracy

C Precision

D Recall

Mean Squared Error (MSE) measures the average squared difference between predicted and actual values, making it suitable for regression error measurement.

What happens if you choose a polynomial degree that is too high?

A The model becomes a linear regression

B The model will always perform better

C The model ignores nonlinear patterns

D The model may overfit the training data

A very high polynomial degree can cause overfitting, where the model fits noise in the training data and performs poorly on new data.

Why is using a pipeline helpful in polynomial regression?

A It visualizes the data

B It automatically tunes hyperparameters

C It combines feature transformation and model fitting steps

D It splits data into batches

A pipeline combines steps like polynomial feature creation and regression fitting into one process, making the workflow simpler and less error-prone.

Which of these is NOT a step in a polynomial regression pipeline?

A Data encryption

B Linear regression fitting

C Polynomial feature transformation

D Model evaluation

Data encryption is unrelated to polynomial regression pipelines, which focus on feature transformation, model fitting, and evaluation.

Explain how a polynomial regression pipeline works from raw data to predictions.

Think about the steps you take to prepare data, train the model, and check results.

Describe the risks and benefits of increasing the polynomial degree in regression.

Consider what happens when the model becomes too simple or too complex.