The IIT Madras BS Machine Learning Practice (MLP) End Term paper sat on 7 Aug 2022, in the May 2022 term: 40 questions for 100 marks in 180 minutes. Every question is below with its answer. Take it as a timed mock test to be marked, or read it through first.
The similarity between loaders and fetchers of sklearn is -
Both loaders and fetchers generate controlled synthetic datasets
Both loaders and fetchers return a Bunch object, which is a dictionary with two keys
Both techniques load small standard datasets
None of these options are correct
Correct answer
Both loaders and fetchers return a Bunch object, which is a dictionary with two keys
Which of the following options represent the main purpose of using the FeatureUnion?
Enables different transformations on various columns of data based on their types
Combines output from several transformer objects by creating a new transformer from them
To chain multiple estimators to execute a fixed sequence of steps in data preprocessing and modeling
None of these
Correct answer
Combines output from several transformer objects by creating a new transformer from them
What will be the output of the following block of code?
(10, 2)
(2, 10)
(10, 3)
(3, 10)
Correct answer
(10, 2)
Which of the following options is the correct method to load the wine dataset and print its shape?
Correct answer
What is the output of the following block of code?
Correct answer
What is the output of the following code snippet?
[0.75 0.5]
[0.25 0.5]
[0.5 0.75]
[0.5 0.5]
Correct answer
[0.5 0.75]
Which of the options will be the correct output for the following code snippet?
import numpy as npx = np.array( [[5, 4 ], [3, 8 ], [2, 0 ], [9, 6 ]])from sklearn.preprocessing import add_dummy_featurex_new = add_dummy_feature(x)print(x_new)Correct answer
Which of the options will be the correct output for the following code snippet?
Correct answer
Which of the following options is correct explained variance score for the following code snippet applying PCA technique.
[1. 1.]
[1. 0.]
[0. 1.]
[0. 0.]
Correct answer
[1. 0.]
Which of the following options is the correct output of the following block of code?
Correct answer
Which of the following options is the correct method to shuffle training data after each epoch in SGDRegressor?
Correct answer
Which of the following options represent the correct output of the following block of code?
array([9.])
array([10.])
array([11.])
SyntaxError: unexpected EOF while parsing
Correct answer
array([10.])
Correct answer
Which of the following options is the possible output for the following block of code?
0.94
1.25
1.94
-2.64
Correct answer
0.94
What operation does the following code implement?
It will perform linear regression on the given data.
It will generate synthetic regression data.
It will generate synthetic classification data.
It will perform classification on the given data.
Correct answer
It will perform classification on the given data.
Which of the following options represent the recall value for the class 2 in the confusion matrix shown in Figure 1 if y_true = [0, 1, 2, 2, 2], y_pred = [0, 0, 2, 2, 1] ?
0.67
1.00
0.50
0.25
Correct answer
0.67
Consider the following code block and mark the correct output.
[0 0 0 1 1 1]
[2 2 0 0 1 1]
[0 1 2 0 1 2]
[0 1 0 1 0 1]
Correct answer
[2 2 0 0 1 1]
Consider X and y as the training dataset. What will be the output of the following code?
logistic
relu
identity
softmax
Correct answer
softmax
Which of the following options represent the design philosophy of the sklearn API?
Nonproliferation of classes
Execution of all codes within 99.9 ms.
Sensible defaults
Direct accessibility of hyperparameters of all estimators
Correct answers
Nonproliferation of classes
Sensible defaults
Direct accessibility of hyperparameters of all estimators
Which of the following options represent(s) the correct method(s) to search for the best regularization parameter for ridge regression?
Search for the best regularization rate with built-in cross validation in RidgeCV estimator
Step 1: Instantiate object of Ridge estimator.
Step 2: Set parameter alpha to the maximum regularization rate.
Use cross validation with Ridge to search for best regularization.
Apply cross validation with SGDRegressor for searching the best regularization parameter.
Correct answers
Search for the best regularization rate with built-in cross validation in RidgeCV estimator
Use cross validation with Ridge to search for best regularization.
Apply cross validation with SGDRegressor for searching the best regularization parameter.
Consider the following block of code and mark all possible correct options.
from sklearn.linear_model import LogisticRegressionfrom sklearn.ensemble import RandomForestClassifier, VotingClassifierclf1 = LogisticRegression()clf2 = RandomForestClassifier()vote=VotingClassifier(estimators=[('lr', clf1), ('rf', clf2)], weights=None, n_jobs= 1, flatten_transform=True, verbose=False, voting='hard')vote = vote.fit(X, y)print(vote.predict(X))Given model uses majority voting rule to predict class labels.
Given model predicts the class labels based on the argmax of the sums of the predicted probabilities.
Setting flatten_transform =True with voting=’soft’ will flatten the output shape of transform.
None of these are correct
Correct answers
Given model uses majority voting rule to predict class labels.
Setting flatten_transform =True with voting=’soft’ will flatten the output shape of transform.
What will be the output of the following code block:
Correct answer
Which of the following options represent the output of the following block of code?
Correct answer
Which of the following statements will be true based on the following diagram?
The classifier will predict all positive classes as negatives and all negative classes as positives.
The classifier will perfectly distinguish between the positive and negative classes.
There is a good probability that the classifier can properly identify the classes.
We can not comment on the classifier performance based on the given figure
Correct answer
The classifier will perfectly distinguish between the positive and negative classes.
Which of the following options is the correct output of the following block of code?
array([2.25, 2.25, 2.25])
array([1., 1., 1.])
array([2., 2., 2.])
DummyRegressor()
Correct answer
array([2., 2., 2.])
Which of the following is likely to be the correct output of the code given below?
array([[1., 1., 1., 0.], [1., 1., 1., 0.], [0., 1., 0., 1.], [0., 1., 1., 1.]])
array([[1., 1., 1., 0.], [1., 1., 1., 0.], [0., 1., 0., 1.], [0., 1., 0., 1.]])
array([[1., 1., 0., 1.], [0., 1., 1., 0.], [0., 1., 1., 0.], [1., 0., 1., 1.]])
array([[1., 1., 1., 0.], [1., 1., 1., 0.], [0., 1., 1., 1.], [0., 1., 1., 1.]])
Correct answer
array([[1., 1., 1., 0.], [1., 1., 1., 0.], [0., 1., 1., 1.], [0., 1., 1., 1.]])
For the following block of code, we get the output as 0.9875. How would the output change if we decrease max depth value ?:
from sklearn.datasets import load_winefrom sklearn.tree import DecisionTreeClassifierfrom sklearn.model_selection import train_test_splitX,y = load_wine(as_frame = True, return_X_y = True)
X_train,X_test,y_train,y_test = train_test_split(X, y, test_size = 0.10, random_state = 12)
clf = DecisionTreeClassifier(max_depth = 6, min_samples_split = 2, min_samples_leaf=3, random_state = 81)
clf.fit(X_train, y_train)print(clf.score(X_train, y_train))Output score will definitely increase.
Output score will definitely decrease.
Output score may decrease or remain the same.
Code will throw an error because max_depth can’t be less than 6.
Correct answer
Output score may decrease or remain the same.
Consider the following block of code and mark the correct option.
from sklearn.datasets import load_breast_cancerfrom sklearn.tree import DecisionTreeClassifierfrom sklearn.model_selection import train_test_splitX,y = load_breast_cancer(as_frame = True, return_X_y = True)X_train,X_test,y_train,y_test = train_test_split(X, y, test_size = 0.2, random_state = 1)clf = DecisionTreeClassifier(min_samples_split = 5, random_state = 5)clf.fit(X_train, y_train)print(clf.score(X_test, y_test))The minimum number of samples required to split a leaf node is 5.
The minimum number of samples required to split an internal node is 6.
The minimum number of samples required to be at a leaf node is 4.
The minimum number of samples required to split an internal node is 5.
Correct answer
The minimum number of samples required to split an internal node is 5.
Consider the following code snippet. What value should we fill at the blank place, so that given code block creates graph as shown below:
2
4
5
6
10
Correct answer
5
Calculate the coefficient of determination (R²) [up to 1 decimal point] based on the following block of code.
Correct answer: 0 (accepted within ±0.05)
In the given block of code, the Iris dataset having shape (150,4) has been loaded for model training. What do you think will be the output of the code given below?
from sklearn.datasets import load_irisfrom sklearn.tree import DecisionTreeClassifierfrom sklearn.model_selection import train_test_splitX,y = load_iris(as_frame = True, return_X_y = True)
X_train,X_test,y_train,y_test = train_test_split(X, y, test_size=0.2, random_state=1)
clf = DecisionTreeClassifier(max_features=2, max_depth = 6, min_samples_split = 2, min_samples_leaf = 3, random_state = 81 )
clf.fit(X_train, y_train)print(clf.n_features_in_)Correct answer: 4
Code snippet written below plots a graph for some dataset X as shown below.
In the given dendrogram which sample is most similar to the sample having label as 1 ?
Correct answer: 0
Go through the code snippet given below and answer the subquestions.
X = [[0,1],[3,4], [4,5],[1,2],[2,3], [5,6]]y = [0, 1, 1,0,0,1]from sklearn.neighbors import RadiusNeighborsClassifierneigh = RadiusNeighborsClassifier(radius=1.0, P=1)neigh.fit(X, y)print(neigh.predict([[2.8,3.8]]))Which of the following is likely to be the correct output of the given code ?
[2]
[0]
[1]
[3]
Correct answer
[1]
Go through the code snippet given below and answer the subquestions.
X = [[0,1],[3,4], [4,5],[1,2],[2,3], [5,6]]y = [0, 1, 1,0,0,1]from sklearn.neighbors import RadiusNeighborsClassifierneigh = RadiusNeighborsClassifier(radius=1.0, P=1)neigh.fit(X, y)print(neigh.predict([[2.8,3.8]]))Which of the following distance metrics is used in the given code.
Euclidean_distance
Manhattan_distance
Hamming Distance
None of these
Correct answer
Euclidean_distance
Consider the following code snippet:
from sklearn.datasets import fetch_california_housingfrom sklearn.neural_network import MLPRegressor
X,y = fetch_california_housing(as_frame = True, return_X_y = True)rs = MLPRegressor(activation='logistic', hidden_layer_sizes=(12,15,13,11,12,8), random_state=12)rs.fit(X, y)Based on the above data, answer the given subquestions.
What will be the output of the following code?
Correct answer: 8
Consider the following code snippet:
from sklearn.datasets import fetch_california_housingfrom sklearn.neural_network import MLPRegressor
X,y = fetch_california_housing(as_frame = True, return_X_y = True)rs = MLPRegressor(activation='logistic', hidden_layer_sizes=(12,15,13,11,12,8), random_state=12)rs.fit(X, y)Based on the above data, answer the given subquestions.
What will be the output of the following code?
(5, 12)
(6, 11)
(8, 15)
(12, 8)
Correct answer
(12, 8)
Go through the code snippet given below and answer the subquestions.
import numpy as npfrom sklearn.pipeline import make_pipelinefrom sklearn.preprocessing import StandardScalerX = np.array([[-8, -3], [3, 3], [5, 3], [-4, -3],])y = np.array([-3, 3, 3, -3])from sklearn.svm import SVCclf = SVC(gamma='auto',kernel="linear")clf.fit(X, y)print(clf.predict([[-4, -3]]))Which of the following is likely to be the correct output of the given code ?
[3]
[-1]
[1]
[-3]
Correct answer
[-3]
Go through the code snippet given below and answer the subquestions.
import numpy as npfrom sklearn.pipeline import make_pipelinefrom sklearn.preprocessing import StandardScalerX = np.array([[-8, -3], [3, 3], [5, 3], [-4, -3],])y = np.array([-3, 3, 3, -3])from sklearn.svm import SVCclf = SVC(gamma='auto',kernel="linear")clf.fit(X, y)print(clf.predict([[-4, -3]]))Which of the following is likely to be the correct output:
array([[-8., -3.], [ 3., 3.]])
array([[-4., -3.], [ 3., 3.]])
array([[-4., -3.], [ 5., 3.]])
array([[-8., -3.], [ 5., 3.]])
Correct answer
array([[-4., -3.], [ 3., 3.]])
Which of the following is/are true about the following block of the code?
from sklearn.neural_network import MLPClassifierX = [[0., 0.], [1., 1.],[2., 3.],[5., 4.]]y = [0, 0, 1, 1]rs = MLPClassifier(solver='adam', alpha=0.0001 , batch_size="auto", hidden_layer_sizes=(8,12), random_state=12)rs.fit(X,y)rs.score(X,y)Given dataset belongs to binary classification.
Number of neurons in the 1^(st) hidden layer is 8.
Strength of the L1 regularization term is 0.0001
The activation function for the output layer in the given code is the ′logistic′ function.
Correct answers
Given dataset belongs to binary classification.
Number of neurons in the 1^(st) hidden layer is 8.
The activation function for the output layer in the given code is the ′logistic′ function.
Which of the following code blocks will throw an error ?
Correct answers