Machine Learning Practice End Term: 21 December 2025 (September 2025 term)

1. A

   Whether a time series has seasonal patterns

2. B

   Whether a time series is stationary

3. C

   Whether the data follows a normal distribution

4. D

   Whether the variance of the data changes over time

Identify the errors in the following code:

1. A

   include_bias has to be kept true for polynomial features.

2. B

   MinMaxScalar should be used for simple LinearRegression Models

3. C

   model.fit_transform should be replace with model.fit

4. D

   model.predict should be replaced with model.predict_proba

What will typically happen if series_with_nans contains NaNs (null values) ?

1. A

   statsmodels will automatically impute NaNs and fit silently.

2. B

   fit() will raise an error or drop NaNs depending on the estimator; user should handle NaNs beforehand.

3. C

   NaNs are treated as zeros.

4. D

   The model will only use the NaN values to estimate variance.

You are using a ColumnTransformer to preprocess a dataset before training a model:

What will be the shape of the transformed matrix X?

1. A

   (3,2)

2. B

   (3,3)

3. C

   (3,4)

4. D

   (4,3)

1. A

   User-based Collaborative Filtering

2. B

   Item-based Collaborative Filtering

3. C

   Content-based filtering

4. D

   Matrix factorization without regularization

We use classification report to evaluate our classification model:

Which of the following metrics are included in the classification report?

Select all that apply.

1. A

   Accuracy

2. B

   Mean Squared Error

3. C

   F1 Score

4. D

   Cross Entropy Loss

5. E

   Recall

1. A

   Increase sparsity

2. B

   Reduce dimensionality

3. C

   Delete user history

4. D

   Improve data imbalance

Consider the following cases for splitting train and test data for some real world dataset.

Which of the following hold true?

1. A

   The selected samples(rows) in X_train will always be same for all the cases.

2. B

   The selected samples(rows) in X_train will always be different for all the cases.

3. C

   The selected samples(rows) in X_train will always be same for case 1 and case 3.

4. D

   The selected samples(rows) in X_train will always be same for case 2 and case 3.

1. A

   Negative Mean Absolute Error

2. B

   Silhoutte Score

3. C

   F1 Score

4. D

   R2 Score

5. E

   Inertia

You are given the following DataFrame:

You want to perform the following imputation strategy: • Age → Replace missing values with the mean Age of each Department • Salary → Replace missing values with the median Salary of the entire column • Department → No imputation Which of the following code snippets correctly performs the above imputation?

1. A

   

2. B

   

3. C

   

4. D

   

1. A

   LinearRegression

2. B

   KNN

3. C

   KMeans

4. D

   DecisionTree

Select all that apply.

1. A

   best_score parameter of grid search returns the mean cross validated score of all the cross validation folds.

2. B

   best_score parameter of grid search returns the max cross validated score of all the cross validation folds.

3. C

   If there are 2 parameter sets A and B with A having 3 values and B having 4 values the total parameter combinations grid search will try is 7.

4. D

   If there are 2 parameter sets A and B with A having 3 values and B having 4 values the total parameter combinations grid search will try is 12.

5. E

   No other scoring metric other than R2 score and mean_squared_error can be used as the scoring metric for GridSearchCV in case of regression problems.

What does metric in the following code measure?

1. A

   Accuracy

2. B

   Precision

3. C

   Recall

4. D

   F1_Score

You are using make_blobs to generate synthetic data for clustering:

What will the output of the above clustering code?

1. A

   One cluster containing all points.

2. B

   It will raise an error since n_samples is not divisible by centers

3. C

   Two clusters, but not necessarily equal-sized; distribution of samples per cluster may vary.

4. D

   Two clusters with exactly 4 points each.

Identify the datatypes of the two datasets given below: (Assume all library imports are correctly done)

1. A

   bunch , bunch

2. B

   bunch , dataframe

3. C

   dataframe , bunch

4. D

   series, dataframe

What does metric in the following code measure?

1. A

   Accuracy

2. B

   Precision

3. C

   Recall

4. D

   F1_Score

1. A

   It starts with all data points in a single cluster and recursively splits them.

2. B

   It requires the number of clusters to be specified before training.

3. C

   It repeatedly merges the two closest clusters until a stopping condition is met.

4. D

   It updates cluster centroids after each merge.

You are training a RandomForestRegressor on a dataset with 20 features and 10,000 samples. Consider the following code:

Which of the following statements is TRUE about how this Random Forest model will behave?

1. A

   With bootstrap=True, every tree sees the entire dataset, but the samples are shuffled randomly for each tree.

2. B

   The forest will use all 20 features for each split, but only 5 trees out of the 200 will be trained on bootstrap samples.

3. C

   Setting max_features=5 means only 5 trees out of 200 will use feature subsets, while the rest use all features.

4. D

   Each tree will consider only 5 randomly chosen features when looking for the best split at every node.

You are using a ColumnTransformer to preprocess a dataset before training a model:

What will be the shape of the transformed matrix X?

1. A

   (3,2)

2. B

   (4,4)

3. C

   (3,4)

4. D

   (4,3)

If pval = 0.03 for the following code, which of the option is correct at the 5% level?

1. A

   Series is non-stationary.

2. B

   Series is stationary.

3. C

   The series has trend.

4. D

   The series is noise.

1. A

   

2. B

   

3. C

   

4. D

   

Select all that apply.

1. A

   n_estimators – Number of trees in the RandomForestClassifier

2. B

   max_depth – Maximum depth allowed for each decision tree

3. C

   min_samples_leaf – Controls the maximum number of trees that will be created

4. D

   bootstrap – Sets the number of features used for splitting at each node

Select all that apply.

1. A

   Negative Mean Absolute Error

2. B

   Silhoutte Score

3. C

   F1 Score

4. D

   R2 Score

5. E

   Inertia

What is the output of the following code?

1. A

   

2. B

   

3. C

   

4. D

   

Select all that apply.

1. A

   Learning the weights of logistic regression model on training data - fit_transform

2. B

   Creating a decision tree on the training data - fit

3. C

   fitting MinMaxScaler on test data - fit_transform

4. D

   fitting MinMaxScaler on train data - fit_transform

Consider the following code snippet. Assume that all required libraries have been imported correctly.

A practitioner observes that each time the fit() method is called on clf, the model is trained from scratch and the parameters learned in the previous training call are discarded. Which of the following changes should be made so that the model continues training from the previously learned parameters when fit() is called again?

1. A

   Set warm_start=True

2. B

   Combine training data from all training sessions into a single dataset

3. C

   Set retain_parameters=True

4. D

   This behavior cannot be changed for Perceptron

Consider the following cases for splitting train and test data for some real world dataset.

Which of the following hold true?

1. A

   The selected samples(rows) in X_train will always be same for all the cases.

2. B

   The selected samples(rows) in X_train will always be different for all the cases.

3. C

   The selected samples(rows) in X_train will always be different for case 1 and case 2.

4. D

   The selected samples(rows) in X_train will always be different for case 1 and case 3.

Following is the code to tune the degree parameter of a polynomial regression model.

What should the blank space contain?

1. A

   "degree"

2. B

   "PolynomialFeatures_degree"

3. C

   "sgd__degree"

4. D

   "poly__degree"

Select all that apply.

1. A

   intercept_ in LinearRegression

2. B

   degree in PolynomialFeatures

3. C

   n_neighbors in Ridge

4. D

   min_samples_split in DecisionTreeClassifier

You are using make_blobs to generate synthetic data for clustering:

What will the output of the above clustering show?

1. A

   One cluster containing all points.

2. B

   It will raise an error since n_samples is not divisible by centers

3. C

   Two clusters, but not necessarily equal-sized; distribution of samples per cluster may vary.

4. D

   Two clusters with exactly 4 points each.

1. A

   

2. B

   

3. C

   

4. D

   

What does the following pandas code do ?

1. A

   All rows where Age < 18, sorted by Marks, index unchanged.

2. B

   Only rows where Marks > 60, sorted by Age, and index reset.

3. C

   All rows sorted first by Marks, then by Age.

4. D

   Drops the Age column and sorts remaining data.

1. A

   50 input features and 30 output units

2. B

   Two hidden layers with 50 and 30 neurons respectively

3. C

   50 epochs and 30 batches

4. D

   50 samples per batch and 30 iterations

1. A

   User–item interaction matrix

2. B

   Item content features

3. C

   User features (e.g. age, gender etc.)

4. D

   Image embeddings

What is the role of the alpha=1.0 parameter in the given sklearn code to build a regression model?

1. A

   It applies L1 regularization and sets some coefficients to zero.

2. B

   It removes regularization completely and behaves like Linear Regression.

3. C

   It adds L2 regularization that penalizes large coefficients to prevent overfitting.

4. D

   It makes the model ignore the intercept term.

Select all that apply.

1. A

   "intercept_" in LinearRegression

2. B

   "degree" in PolynomialFeatures

3. C

   "n_neighbors" in KNeighborsClassifier

4. D

   "max_depth" in DecisionTreeClassifier

Given the following code using BaggingClassifier with KNeighborsClassifier as the base estimator:

Which of the following statements is correct?

1. A

   Above code uses bootstrapping to generate samples for each base classifier.

2. B

   model will be tested on out of the bags samples.

3. C

   Due to weights="distance", each base KNN classifier will treat all neighbors equally in terms of voting power.

4. D

   The ensemble will consist of 3 base KNN classifiers.

5. E

   None of these

Consider the following code snippet that trains a Decision Tree classifier using scikit-learn:

A node in the trained decision tree is split only if both of the following conditions are satisfied: • The node contains at least min_samples_split samples. • Each child node created after the split contains at least min_samples_leaf samples. Assume node (N) is a non-leaf node being considered for splitting. In which of the following cases will the split at node (N) be allowed?

Select all that apply.

1. A

   Node (N) contains 15 samples. After splitting, the left child has 9 samples and the right child has 6 samples.

2. B

   Node (N) contains 6 samples. After splitting, the left child has 4 samples and the right child has 2 samples.

3. C

   Node (N) contains 12 samples. After splitting, the left child has 3 samples and the right child has 9 samples.

4. D

   Node (N) contains 8 samples. After splitting, the left child has 4 samples and the right child has 4 samples.

Select all that apply.

1. A

   LinearRegression

2. B

   DecisionTree

3. C

   SVM

4. D

   NaiveBayes

Consider the following ensemble model built using VotingClassifier with voting='soft':

How is the final prediction for a test sample obtained?

1. A

   By selecting the most common predicted class.

2. B

   By averaging predicted class probabilities.

3. C

   By summing the predicted class labels.

4. D

   By choosing the most accurate classifier.

Given below a y_train list which consists of coffee order's preference by the customers in a cafe.

MultiLabelBinarizer from sklearn library has been used to convert the y_train into numbers, so which of the following option matches with the output of the following code ?

1. A

   [[1 0 1 0],[1 0 0 1],[0 1 0 1],[0 1 1 0]]

2. B

   [[0 0],[1 0],[0 1],[1 1],[1 0],[0 0],[0 0],[0 0]]

3. C

   [[1 0 1 0 0 1 0 0],[1 0 0 1 1 0 0 0]]

4. D

   [[1 0][1 0],[1 0] [0 1],[0 1] [0 1],[0 1][1 0]]

You are given a dataset where each row represents a customer using two numerical features. You apply K-means clustering with K=3 clusters using the following code:

What does the variable labels represent after the model is fitted?

1. A

   The values of the input features used for clustering.

2. B

   The cluster index assigned to each data point.

3. C

   The total number of clusters formed.

4. D

   The coordinates of the cluster centroids.

What could be the output for below code

1. A

   ['document' 'first' 'is' 'second' 'the' 'this']

2. B

   {'this': 5, 'is': 2, 'the': 4, 'first': 1, 'document': 0, 'second': 3}

3. C

   [5, 2, 4, 1, 0, 3]

4. D