👉1. Machine Learning MCQs - UNIT 1
👉2. Machine Learning MCQs-2 (Performance Metrics, Linear Regression, Naïve Bayes Classifier )
👉3. Machine Learning MCQs-3 (Logistic Regression, KNN, SVM, Decision Tree)
👉4. Machine Learning MCQs - 4 (Clustering, Dimensionality Reduction)
MACHINE LEARNING Syllabus:
UNIT-1
Introduction: Brief Introduction to Machine Learning, Abstraction and Knowledge
Representation, Types of Machine Learning Algorithms, Definition of learning
systems, Goals and applications of machine learning, Aspects of developing a learning
system, Data Types, training data, concept representation, function
approximation.
Data
Pre-processing:
Definition, Steps involved in
pre-processing, Techniques
UNIT-2
Performance
measurement of models: Accuracy, Confusion
matrix, TPR, FPR, FNR, TNR, Precision, recall, F1-score, Receiver Operating
Characteristic Curve (ROC) curve and AUC.
Supervised
Learning1:
Linear Regression, Multiple Variable Linear
Regression, Naïve Bayes Classifiers, Gradient Descent, Multicollinearity,
Bias-Variance trade-off.
UNIT-3
Supervised
Learning2:
Regularization, Logistic Regression,
Squashing function, KNN, Support Vector Machine.
Decision Tree Learning:
Representing concepts as decision trees, Recursive induction of decision trees,
picking the best splitting attribute: entropy and information gain, searching
for simple trees and computational complexity, Occam's razor, overfitting,
noisy data, and pruning. Decision Trees – ID3-CART-Error bounds.
UNIT-4
Unsupervised
Learning:
K-Means, Customer Segmentation,
Hierarchical clustering, DBSCAN, Anomaly Detection, Local Outlier Factor,
Isolation Forest, Dimensionality Reduction, PCA, GMM, Expectation Maximization.
UNIT-5
Ensemble Models: Ensemble Definition, Bootstrapped Aggregation (Bagging) Intuition,
Random Forest and their construction, Extremely randomized trees, Gradient
Boosting, Regularization by Shrinkage, XGBoost, AdaBoost.
TEXT
BOOKS:
1. Machine Learning – Tom M. Mitchell, - MGH
2. Ethem Alpaydin, “Introduction to Machine Learning”, MIT Press,
Prentice Hall of India, Third Edition 2014.
3. The Elements of Statistical Learning, Trevor Hastie, Robert
Tibshirani & Jerome Friedman, Springer Verlag, 2001.
REFERENCES:
-
1.
Machine Learning, SaikatDutt, Subramanian
Chandramouli, Amit Kumar Das, Pearson, 2019.
2.
Stephen Marsland, “Machine Learning -An Algorithmic
Perspective”, Second Edition, Chapman and Hall/CRC Machine Learning and Pattern
Recognition Series, 2014.
3.
Application of machine learning in
industries (IBM ICE Publications).
e-Resources:
1.
Andrew Ng, “Machine Learning Yearning” https://www.deeplearning.ai/machinie-learning
2.
Shai Shalev-Shwartz, Shai Ben-David,
“Understanding Machine Learning: From Theory to Algorithms”, Cambridge
University Press.
https://www.cs.huji.ac.il/w~shais/UnderstaningMachineLearning/index.html
1. The model which consists of a
group of predictors is called a
Ans: C
2.
A Random forest is an ensemble of Decision Trees
Ans: A
3.
The steps involved in deciding the output of a Random Forest are
Ans: C
4.
A hard voting classifier takes into consideration
Ans: B
5. If each classifier is a weak learner, the ensemble can still be a strong
learner?
Ans: A
6. Ensemble methods work best when the predictors are
Ans: D
7. To get
diverse classifiers we cannot train them using different algorithms
Ans: B
8. Training
the classifiers in an ensemble using very different algorithms increases the
chance that they will make very different types of errors, improving the
ensemble’s accuracy
Ans: A
9. When we consider only the majority of the outputs
from the classifiers then it is called
Ans: A
10. Soft voting takes into consideration
Ans: B
11. In soft voting, the predicted class is the class
with the highest class probability, averaged over all the individual
classifiers
Ans: A
12. Soft voting achieves higher performance than hard
voting because
Ans: B
13. The parameter which decides the voting method in
a VotingClassifier is
Ans: C
14. The parameter which holds the list of classifiers
which are to be used in the voting classifier is
Ans: C
15. One way to get a diverse set of classifiers is to
use the same training algorithm for every predictor, but to train them on
different random subsets of the training set
Ans: A
16. When sampling is performed with replacement, the
method is
A. Bagging
B. Pasting
C. Both
D. None
Ans: A
17. When sampling is performed without replacement,
it is called
Ans: A
18. Both bagging and pasting allow training instances
to be sampled several times across multiple predictors, but only bagging allows
training instances to be sampled several times for the same predictor
Ans: A
19. In
bagging/pasting training set sampling and training, predictors can all be
trained in parallel, via different CPU cores or even different servers
Ans: A
20. To use
the bagging method, the value of the bootstrap parameter in the
BaggingClassifier should be set to
Ans: A
21. To use the pasting method, the value of the
bootstrap parameter in the BaggingClassifier should be set to
A. True
B. False
Ans: B
22. Overall, bagging often results in better models
Ans: A
23. How many training instances with replacement does
the BaggingClassifier train if the size of the training set is m?
Ans: C
24. With bagging, it is not possible that some
instances are never sampled
Ans: B
25. Features can also be sampled in the
BaggingClassifier
Ans: A
26. The hyperparameters which control the feature
sampling are
Ans: B
27. Sampling both training instances and features is
called the
Ans: A
28. Keeping all training instances (i.e.,
bootstrap=False and max_samples=1.0) but sampling features (i.e.,
bootstrap_features=True and/or max_features smaller than 1.0) is called the
Ans: B
29. Random
forest is an ensemble of Decision Trees generally trained via ______
Ans: A
30. We can make the trees of a Random Forest even
more random by using random thresholds for each feature rather than searching
for the best possible thresholds?
Ans: B
31. If we look at a single Decision Tree, important
features are likely to appear closer to
Ans: C
32. Feature importances are available via the
feature_importances_ method of the RandomForestClassifier object.
Ans: A
33. The general idea of most boosting methods is to
train predictors sequentially, each trying to correct its predecessor.
Ans: A
34. One of the drawbacks of AdaBoost classifier is
that
Ans: B
35. A
Decision Stump is a Decision Tree with
Ans: B
36. In
Gradient Boosting, instead of tweaking the instance weights at every iteration
like AdaBoost does, it tries to fit the new predictor to the residual errors
made by the previous predictor.
Ans: A
37. The learning_rate hyperparameter of
GradientBoostingRegressor scales the contribution of each tree?
Ans: A
38. The
ensemble method in which we train a model to perform the aggregation of outputs
from all the predictors is called
Ans: C
