Back to Subject

Unit 4: Regression with scikit-Learn - Practice Quiz

INT395 — Supervised Learning 50 Questions
0 Correct 0 Wrong 50 Left
0/50

1 In the context of supervised learning, what distinguishes a regression problem from a classification problem?

A. The input features must be categorical.
B. The target variable is categorical.
C. The target variable is continuous.
D. Regression requires unsupervised data.

2 Which visualization tool is most commonly used in Exploratory Data Analysis (EDA) to visualize the linear relationship between a single feature and the target variable?

A. Box plot
B. Histogram
C. Pie chart
D. Scatter plot

3 When analyzing the relationship between multiple variables in a dataset, which matrix helps quantify the linear correlation between every pair of features?

A. Correlation matrix
B. Covariance matrix
C. Confusion matrix
D. Hessian matrix

4 In a simple linear regression model , what does represent?

A. The y-intercept
B. The slope of the line
C. The residual error
D. The learning rate

5 Which Scikit-Learn class is used to perform Ordinary Least Squares (OLS) linear regression?

A. sklearn.linear_model.SGDRegressor
B. sklearn.linear_model.Lasso
C. sklearn.linear_model.Ridge
D. sklearn.linear_model.LinearRegression

6 What is the objective function that Ordinary Least Squares (OLS) minimizes?

A. Mean Absolute Error (MAE)
B. Sum of Squared Errors (SSE)
C. Cross-Entropy Loss
D. Hinge Loss

7 What is the primary motivation for using the RANSAC (RANdom SAmple Consensus) algorithm in regression?

A. To increase the speed of training.
B. To fit a model in the presence of a significant number of outliers.
C. To perform feature selection.
D. To handle missing values automatically.

8 In the RANSAC algorithm, what does the 'residual_threshold' parameter define?

A. The maximum number of iterations.
B. The learning rate of the estimator.
C. The minimum number of samples required to fit the model.
D. The maximum residual for a data sample to be classified as an inlier.

9 Which metric is calculated as ?

A. Mean Squared Error
B. Coefficient of Determination
C. Explained Variance Score
D. Mean Absolute Error

10 If an score is 1.0, what does this indicate about the regression model?

A. The model explains none of the variability of the response data.
B. The model is a constant line.
C. The model perfectly fits the data.
D. The model is underfitting.

11 Why is Mean Squared Error (MSE) often preferred over Mean Absolute Error (MAE) for optimization?

A. MSE is robust to outliers.
B. MSE is always smaller than MAE.
C. MSE is differentiable everywhere, making gradient-based optimization easier.
D. MSE has the same unit as the target variable.

12 To perform Polynomial Regression using a linear model in Scikit-Learn, which transformer must be applied first?

A. PolynomialFeatures
B. SimpleImputer
C. OneHotEncoder
D. StandardScaler

13 What is the main risk associated with using a high-degree polynomial in regression?

A. Convergence failure
B. High bias
C. Underfitting
D. Overfitting

14 Which of the following techniques helps reduce overfitting in regression models by adding a penalty term to the loss function?

A. Standardization
B. Augmentation
C. Normalization
D. Regularization

15 Ridge regression minimizes the sum of squared residuals plus a penalty term based on:

A. The sum of absolute values of coefficients ( norm).
B. The sum of squared values of coefficients ( norm).
C. The number of non-zero coefficients.
D. The maximum coefficient value.

16 Which property makes Lasso regression useful for feature selection?

A. It forces some coefficients to become exactly zero.
B. It works best when .
C. It shrinks coefficients uniformly.
D. It increases the magnitude of coefficients.

17 In Scikit-Learn, which regression model combines both and regularization penalties?

A. Lasso
B. BayesianRidge
C. Ridge
D. ElasticNet

18 In regularized regression, what is the role of the hyperparameter (or )?

A. It determines the degree of the polynomial.
B. It controls the strength of the regularization penalty.
C. It controls the learning rate.
D. It sets the intercept to zero.

19 Why is feature scaling (e.g., Standardization) important before applying Ridge or Lasso regression?

A. To convert categorical data to numeric.
B. To ensure the target variable is normally distributed.
C. To remove missing values.
D. Because the penalty term is sensitive to the scale of the coefficients.

20 Support Vector Regression (SVR) tries to fit as many data points as possible within a margin of width:

A.
B. (epsilon)
C.
D. Zero

21 In Support Vector Regression, what is the role of the kernel function?

A. To map the input data into a higher-dimensional feature space to handle non-linearity.
B. To normalize the target variable.
C. To calculate the error metric.
D. To select the best features.

22 Which parameter in SVR controls the trade-off between the smoothness of the decision function and the tolerance for training errors?

A. C
B. Gamma
C. Kernel
D. Degree

23 What is the primary criterion used by Decision Tree Regressors to split a node?

A. Log-Loss
B. MSE (Variance reduction)
C. Information Gain
D. Gini Impurity

24 One major advantage of Decision Tree Regression is:

A. It never overfits.
B. It creates smooth, continuous prediction curves.
C. It does not require feature scaling or normalization.
D. It always extrapolates well.

25 What is a characteristic behavior of a Decision Tree Regressor when predicting values outside the range of the training data?

A. It predicts the average of the closest training samples (constant prediction).
B. It extrapolates linearly.
C. It returns a null value.
D. It automatically creates a polynomial fit.

26 Random Forest Regression improves upon a single Decision Tree by utilizing which technique?

A. Pruning
B. Bagging (Bootstrap Aggregating)
C. Kernel trick
D. Gradient Boosting

27 In a Random Forest Regressor, how is the final prediction determined?

A. Weighted sum of the features.
B. Average of the predictions of all individual trees.
C. Majority vote of the trees.
D. The prediction of the tree with the highest accuracy.

28 Which parameter in RandomForestRegressor determines the number of trees in the forest?

A. n_estimators
B. bootstrap
C. max_depth
D. min_samples_split

29 Random Forests introduce randomness in two ways: bootstrap sampling and:

A. Randomly pruning the trees.
B. Randomly shuffling the target labels.
C. Selecting a random subset of features at each split.
D. Random initialization of weights.

30 What is the 'Out-of-Bag' (OOB) score in Random Forests?

A. The error rate of the worst tree.
B. A validation score calculated using the samples not included in the bootstrap sample for each tree.
C. The accuracy on the test set.
D. The training error of the full ensemble.

31 Which Scikit-Learn function splits a dataset into training and testing sets?

A. sklearn.preprocessing.train_test
B. sklearn.model_selection.train_test_split
C. sklearn.model_selection.cross_val_score
D. sklearn.metrics.split_data

32 In the equation for ElasticNet: , what does (or l1_ratio in scikit-learn) control?

A. The tolerance for stopping criteria.
B. The mix between Ridge and Lasso regularization.
C. The overall regularization strength.
D. The degree of the polynomial.

33 A residual plot shows the residuals on the y-axis and the predicted values on the x-axis. What pattern indicates a good regression model?

A. A clear U-shape curve.
B. A funnel shape (heteroscedasticity).
C. Points randomly scattered around the horizontal axis (zero).
D. A linear trend.

34 When using SGDRegressor from Scikit-Learn, which hyperparameter defines the update rule schedule (how the learning rate changes over time)?

A. alpha
B. learning_rate
C. penalty
D. loss

35 Which of the following is an intrinsic weakness of Linear Regression?

A. It is computationally expensive.
B. It is difficult to interpret.
C. It cannot model non-linear relationships without feature engineering.
D. It requires categorical features.

36 In the context of regression metrics, what does Median Absolute Error provide that Mean Absolute Error does not?

A. Percentage error calculation.
B. Squared penalization.
C. Differentiability.
D. Robustness to outliers.

37 What is Multicollinearity?

A. When the model has too many polynomial features.
B. When the training data is too small.
C. When the target variable is categorical.
D. When independent features are highly correlated with each other.

38 How does DecisionTreeRegressor handle missing values in Scikit-Learn (standard implementation)?

A. It handles them natively.
B. It ignores the rows with missing values.
C. It requires imputation (filling missing values) before training.
D. It treats them as a separate category.

39 Which plot is typically used to inspect if the residuals follow a normal distribution?

A. Box plot
B. Scatter plot
C. Q-Q (Quantile-Quantile) plot
D. Bar chart

40 In Polynomial Regression, if you increase the degree of the polynomial significantly, the model becomes:

A. More biased.
B. Linear.
C. More complex with higher variance.
D. Less flexible.

41 What is the result of applying fit_transform on the training data and then transform on the test data during scaling?

A. Overfitting.
B. Incorrect scaling.
C. Data leakage.
D. Correct application of preprocessing parameters learnt from training to test data.

42 Which Scikit-Learn attribute holds the estimated coefficients for a Linear Regression model after fitting?

A. model.weights_
B. model.coef_
C. model.intercept_
D. model.params_

43 What is the analytical solution to find the optimal weights for Linear Regression called?

A. Coordinate Descent
B. Gradient Descent
C. The Normal Equation
D. Backpropagation

44 When using Support Vector Regression with an RBF kernel, what happens if the parameter (gamma) is very large?

A. The influence of each training example reaches very far.
B. The influence of each training example is limited to a close radius, leading to overfitting.
C. The model behaves like a linear regression.
D. The model becomes a flat line.

45 Why might one use Adjusted instead of standard ?

A. To account for the number of predictors, penalizing the addition of useless features.
B. To handle categorical variables.
C. To ensure the score is always positive.
D. To calculate error in absolute terms.

46 In the context of Bias-Variance tradeoff, a simple linear model with few features typically has:

A. Low Bias and Low Variance
B. High Bias and Low Variance
C. High Bias and High Variance
D. Low Bias and High Variance

47 Which of the following creates a pipeline in Scikit-Learn that scales data then fits a regressor?

A. sklearn.compose.ColumnTransformer()
B. sklearn.pipeline.make_pipeline(StandardScaler(), LinearRegression())
C. sklearn.linear_model.LinearRegression(normalize=True)
D. sklearn.model_selection.cross_val_score(LinearRegression())

48 What is the interpretation of the slope coefficient in the model ?

A. The correlation between and .
B. The percentage change in .
C. The change in for a one-unit increase in .
D. The value of when .

49 Which regression algorithm constructs a model based on the principle of 'recursive binary splitting'?

A. Decision Tree Regression
B. Ridge Regression
C. Support Vector Regression
D. Linear Regression

50 When interpreting a heatmap of a correlation matrix, a value of -0.9 between two features indicates:

A. A strong positive linear relationship.
B. No linear relationship.
C. A strong negative linear relationship.
D. A weak negative linear relationship.
Previous MCQ
Unit 3
Next MCQ
Unit 5