You are on page 1of 31

Data From Jaffar

QUESTION 1
Learning in biological systems involves

a. Addition of new neurons and connections to the system


b. None of the mentioned
c. Adjustments to the synaptic connections that exist between the neurons
Feedback
The correct answer is: Adjustments to the synaptic connections that exist between the neurons

QUESTION 2
An ANN can be characterized by:

a. Architecture, Neurons, Activation Function


b. Architecture, Training or Learning Algorithms, Activation Function
c. Architecture, Training or Learning Algorithms, Neurons
d. Architecture, Neurons, Layers
Feedback
The correct answer is: Architecture, Training or Learning Algorithms, Activation Function

QUESTION 3
Artificial neuron ............... binary input and ................ binary output

a. One, more than 1


b. one or more, one
c. One or more, greater than 1
d. One or more, one or more
Feedback
The correct answer is: one or more, one

QUESTION 4
Which key word is used to pandas in python.

a. None
b. pd
c. Pd
d. Pad
Feedback
The correct answer is: pd

QUESTION 5
In ANNs the information passing between neurons is

a. Fast as compare to the biological counterparts


b. Same as in the biological counterparts
c. None of the mentioned
d. Slow as compare to the biological counterparts
Feedback
The correct answer is: Fast as compare to the biological counterparts

QUESTION 6
Numpy is also consider essential for other packages

a. Pandas\Scikit-learn
b. Scikit-learn
c. Pandas
d. Pandas
Feedback
The correct answer is: Scikit-learn

QUESTION 7
Feedforward neural network the signal is ........... directional

a. Uni
b. Arbitrary
c. All of these
d. Multi
Feedback
The correct answer is: Uni

QUESTION 8
Which package is consider standard for storing numerical data.

a. NumPy\Pandas
b. Pandas
c. Numpy
d. None
Feedback
The correct answer is: Numpy

QUESTION 9
While XOR classification problem output will be 1 if the inputs are:

a. Both B & C
b. (1,0)
c. (0,1)
d. (0,0)
Feedback
The correct answer is: Both B & C

QUESTION 10
In perception algorithm penalty is denoted as:

a. eta0
b. Constant
c. None
d. eta1
Feedback
The correct answer is: eta0

QUESTION 11
Artificial Neural Networks are composed of a large number of highly interconnected processing elements called

a. Blocks
b. Neurons
c. Cells
d. Connections
Feedback
The correct answer is: Neurons

QUESTION 12
While logical computations with neurons, neuron C gets activated if either neuron A or neuron B is activated,
which logic will be used to do so?

a. OR
b. AND
c. NAND
d. NOT
Feedback
The correct answer is: OR

QUESTION 13
If decision boundary is non-linear so the perceptions are ................... of learning complex pattern

a. Incapable
b. None of these
c. Capable
d. Both A & B
Feedback
The correct answer is: Capable

QUESTION 14
When a data scientist needs to deals with two dimentional data in python, than which package is preferable

a. None
b. Scikit-learn
c. Numpy
d. Pandas
Feedback
The correct answer is: Pandas

QUESTION 15
Which one is true about neural networks:

a. In artificial systems, neurons are not connected.


b. In biological systems, neurons are connected much less than in artificial system.
c. In biological systems, neurons are connected much more than in artificial system.
d. In biological systems, neurons are not connected.
Feedback
The correct answer is: In biological systems, neurons are connected much more than in artificial system.

QUESTION 1
 a. False
 b. True
 c. Depends on Data
 d. Depends on Model
Feedback
The correct answer is: True

QUESTION 2
By considering a thumb rule, precision/recall is preferred whenever the ............ class is rare
 a. Positive
 b. Negative
 c. Extrinsic
 d. Intrinsic
Feedback
The correct answer is: Positive

QUESTION 3
Feature engineering involves:
 a. Feature extraction
 b. Feature selection
 c. Creating new features
 d. All of these
Feedback
The correct answer is: All of these

QUESTION 4
Select the commonly best ratio for training the data and holding the data, respectively.
 a. 70% and 30%
 b. 80% and 20%
 c. 40% and 60%
 d. 60% and 40%
Feedback
The correct answer is: 80% and 20%

QUESTION 5
In Online learning small groups are called:
 a. Micro-batches
 b. Nano-batches
 c. None of these
 d. Mini-batches
Feedback
The correct answer is: Mini-batches
QUESTION 6
Choose the odd one.
 a. Support vector machine
 b. Multidimensional Scaling
 c. Principal Component Analysis
 d. Linear Discriminant Analysis
Feedback
The correct answer is: Support vector machine

QUESTION 7
By considering large data which one is preferable?
 a. GridSearchCV
 b. RandomizedSearchCV
 c. SearchSpaceCV
 d. EnsembleSearchCV
Feedback
The correct answer is: RandomizedSearchCV

QUESTION 8
Variables in Python are case-sensitive name.
 a. False
 b. Depends on Editor
 c. True
Feedback
The correct answer is: True

QUESTION 9
A dense NumPy array would be our output when we are using.
 a. Both A & b
 b. LabelBinarize
 c. One-hot encoding
Feedback
The correct answer is: One-hot encoding

QUESTION 10
Locally Linear Embedding (LLE) is a very powerful nonlinear dimensionality reduction technique
 a. False
 b. True
Feedback
The correct answer is: True

QUESTION 11
If I am using all features of my dataset and I achieve 100% accuracy on my training set, but ~70% on validation set,
what should I look out for?
 a. Underfitting
 b. Overfitting
 c. Data Partition
 d. Nothing, the model is perfect
Feedback
The correct answer is: Overfitting

QUESTION 12
Some telecommunication company wants to segment their customers into distinct groups in order to send
appropriate subscription offers, this is an example of
 a. Unsupervised learning
 b. Serration
 c. Data extraction
 d. Supervised learning
Feedback
The correct answer is: Unsupervised learning

QUESTION 13
Support vectors are the data points that lie closest to the decision surface.
 a. TRUE
 b. FALSE
Feedback
The correct answer is: TRUE

QUESTION 14
In an election, N candidates are competing against each other and people are voting for either of the candidates.
Voters don’t communicate with each other while casting their votes. => Now which of the following ensemble
method works similar to above-discussed election procedure? Hint: Persons are like base models of ensemble
method.
 a. Learn++
 b. Bagging
 c. Boosting
 d. AdaBoost
Feedback
The correct answer is: Bagging

QUESTION 15
Distance between the support vectors is called:
 a. Margin
 b. Response
 c. Separator
 d. Hyperplane
Feedback
The correct answer is: Margin

QUESTION 16
The best dimensionality reduction algorithm is called
 a. Isomap
 b. PCA
 c. LDA
 d. LLE
Feedback
The correct answer is: Isomap

QUESTION 17
Classes that have same application programming interface are.
 a. Extra trees regressor and random trees regressor
 b. Sklearn.utils.Bunch and sklearn.utils.close
 c. Description and detail
 d. numpy.ndarray and iris.target
Feedback
The correct answer is: Extra trees regressor and random trees regressor

QUESTION 18
When a classifier is trained, it stores the list of target classes in its classes_ attribute.
 a. In descending order
 b. Randomly
 c. In ascending order
 d. Ordered by value
Feedback
The correct answer is: Ordered by value

QUESTION 19
Confusion Matrix returns
 a. Confused values
 b. Dataset evaluators
 c. Both A & B
 d. Predictions
Feedback
The correct answer is: Predictions

QUESTION 20
Pasting and bagging can be done on same algorithm.
 a. False
 b. True
Feedback
The correct answer is: True

QUESTION 21
The error rate on old cases is called............
 a. Overfitting error
 b. None of these
 c. Normalization error
 d. Generalization error
Feedback
The correct answer is: None of these

QUESTION 22
 a. All of these
 b. Creating new features
 c. Feature extraction
 d. Feature selection
Feedback
The correct answer is: All of these

QUESTION 23
What is true about an ensembled classifier? (1) Classifiers that are more “sure” can vote with more conviction. (2)
Classifiers can be more “sure” about a particular part of the space. (3) Most of the times, it performs better than a
single classifier.
 a. 2 and 3
 b. 1 and 3
 c. All of the above
 d. 1 and 2
Feedback
The correct answer is: All of the above

QUESTION 24
Ensembles will yield bad results when there is significant diversity among the models when all individual models
have meaningful and good predictions.
 a. May or may not
 b. True
 c. False
Feedback
The correct answer is: False
QUESTION 25
 a. Sequence
 b. None of these
 c. a & b
 d. Summary
Feedback
The correct answer is: Summary

QUESTION 26
To compare the classifiers one should have to measure
 a. Both a & b
 b. Area under the curve
 c. The size of the all points
 d. Area left on the right of the result
Feedback
The correct answer is: Area under the curve

QUESTION 27
A multinomial classifier can be used as binomial classier.
 a. Always False
 b. Only if Data is Multi-label
 c. Depends on Data
 d. Always True
Feedback
The correct answer is: Always True

QUESTION 28
>>> X, y = mnist["data"], mnist["target"] >>>, X.shape(70000, 784). Therefore, choose X.shape(?, ?)
 a. Both a & b
 b. X.shape(feature, images)
 c. X.shape(images, feature)
 d. X.shape(MNIST, shape)
Feedback
The correct answer is: X.shape(images, feature)

QUESTION 29
How can SVM be classified?
 a. It is a model trained using unsupervised learning. It can be used for classification but not for regression.
 b. It is a model trained using unsupervised learning. It can be used for classification and regression.
 c. It is a model trained using unsupervised learning. It can be used for classification but not for regression.
 d. It is a model trained using supervised learning. It can be used for classification and regression.
Feedback
The correct answer is: It is a model trained using supervised learning. It can be used for classification and
regression.

QUESTION 30
While creating workspace directory of our Machine Learning code and datasets. The command is written..............
$ prompts.
 a. Both A & B
 b. At start of
 c. After the
 d. Before the
Feedback
The correct answer is: After the

QUESTION 31
 a. 0 to 3
 b. 0 to 4
 c. 1 to 4
 d. 1 to 3
Feedback
The correct answer is: 0 to 4

QUESTION 32
Which of these statement(s) is/are true about k-NN algorithm?. (1) k-NN performs much better if all of the data
have the same scale. (2) k-NN works well with a small number of input variables (p), but struggles when the
number of inputs is very large. (3) k-NN makes no assumptions about the functional form of the problem being
solved
 a. All 1, 2 and 3
 b. Only 1
 c. 1 and 3
 d. 1 and 2
Feedback
The correct answer is: All 1, 2 and 3

QUESTION 33
Binary classifiers cannot be combined to generate a multinomial classier.
 a. Sometimes
 b. Depends on Data
 c. True
 d. False
Feedback
The correct answer is: False

QUESTION 34
Train a binary classifier, capable of distinguishing between just two classes 6 and not-6. Choose a random state.
 a. y_train_5 = (y_train == 6) <= y_test_5 = (y_test == 6)
 b. y_train_5 = (y_train == 6); y_test_5 = (y_test == 6)
 c. y_train_6 = (y_train == 6) && y_test_6 = (y_test != 6)
 d. y_train_5 = (y_train == 6) || y_test_5 = (y_test != 6)
Feedback
The correct answer is: y_train_5 = (y_train == 6); y_test_5 = (y_test == 6)
Data From Ali Faran
Final paper MCQ’s
QUESTION 1
In an election, N candidates are competing against each other and people are voting
for either of the candidates. Voters don’t communicate with each other while casting
their votes. => Now which of the following ensemble method works similar to above-
discussed election procedure? Hint: Persons are like base models of ensemble
method.
a.
Boosting
b.
Bagging
c.
AdaBoost
d.
Learn++
Feedback
The correct answer is: Bagging
QUESTION 2
If I am using all features of my dataset and I achieve 100% accuracy on my training set,
but ~70% on validation set, what should I look out for?
a.
Nothing, the model is perfect
b.
Data Partition
c.
Overfitting
d.
Underfitting
Feedback
The correct answer is: Overfitting
QUESTION 3
Select the commonly best ratio for training the data and holding the data, respectively.
a.
60% and 40%
b.
40% and 60%
c.
80% and 20%
d.
70% and 30%
Feedback
The correct answer is: 80% and 20%
QUESTION 4
When a classifier is trained, it stores the list of target classes in its classes_ attribute.
a.
Ordered by value
b.
In ascending order
c.
Randomly
d.
In descending order
Feedback
The correct answer is: Ordered by value
QUESTION 5
How can SVM be classified?
a.
It is a model trained using unsupervised learning. It can be used for classification but not for
regression.
b.
It is a model trained using supervised learning. It can be used for classification and regression.
c.
It is a model trained using unsupervised learning. It can be used for classification and regression.
d.
It is a model trained using unsupervised learning. It can be used for classification but not for
regression.
Feedback
The correct answer is: It is a model trained using supervised learning. It can be used for
classification and regression.
QUESTION 6
In SGDClassifier If you want reproducible results, you should set the random_state
parameter.
a.
True
b.
False
Feedback
The correct answer is: True
QUESTION 7
Some telecommunication company wants to segment their customers into distinct
groups in order to send appropriate subscription offers, this is an example of
a.
Data extraction
b.
Supervised learning
c.
Serration
d.
Unsupervised learning
Feedback
The correct answer is: Unsupervised learning
QUESTION 8
Train a binary classifier, capable of distinguishing between just two classes 6 and not-6.
Choose a random state.
a.
y_train_5 = (y_train == 6) <= y_test_5 = (y_test == 6)
b.
y_train_5 = (y_train == 6) || y_test_5 = (y_test != 6)
c.
y_train_5 = (y_train == 6); y_test_5 = (y_test == 6)
d.
y_train_6 = (y_train == 6) && y_test_6 = (y_test != 6)
Feedback
The correct answer is: y_train_5 = (y_train == 6); y_test_5 = (y_test == 6)
QUESTION 9
The mean squared distance between the original data and the reconstructed data is
called
a.
Both a & b
b.
Reusability error
c.
Reconstruction error
d.
Rearrange able error
Feedback
The correct answer is: Reconstruction error
QUESTION 10
By considering a thumb rule, precision/recall is preferred whenever the ............ class is
rare
a.
Extrinsic
b.
Intrinsic
c.
Positive
d.
Negative
Feedback
The correct answer is: Positive
QUESTION 11
a.
All of these
b.
Feature extraction
c.
Feature selection
d.
Creating new features
Feedback
The correct answer is: All of these
QUESTION 12
Classes that have same application programming interface are.
a.
Extra trees regressor and random trees regressor
b.
numpy.ndarray and iris.target
c.
Sklearn.utils.Bunch and sklearn.utils.close
d.
Description and detail
Feedback
The correct answer is: Extra trees regressor and random trees regressor
QUESTION 13
Choose the odd one.
a.
Linear Discriminant Analysis
b.
Multidimensional Scaling
c.
Support vector machine
d.
Principal Component Analysis
Feedback
The correct answer is: Support vector machine
QUESTION 14
What is true about an ensembled classifier? (1) Classifiers that are more “sure” can
vote with more conviction. (2) Classifiers can be more “sure” about a particular part of
the space. (3) Most of the times, it performs better than a single classifier.
a.
1 and 2
b.
All of the above
c.
2 and 3
d.
1 and 3
Feedback
The correct answer is: All of the above
QUESTION 15
Confusion Matrix returns
a.
Confused values
b.
Both A & B
c.
Dataset evaluators
d.
Predictions
Feedback
The correct answer is: Predictions
QUESTION 16
a.
1 to 3
b.
1 to 4
c.
0 to 4
d.
0 to 3
Feedback
The correct answer is: 0 to 4
QUESTION 17
In Online learning small groups are called:
a.
Micro-batches
b.
Mini-batches
c.
Nano-batches
d.
None of these
Feedback
The correct answer is: Mini-batches
QUESTION 18
The best dimensionality reduction algorithm is called
a.
PCA
b.
LDA
c.
Isomap
d.
LLE
Feedback
The correct answer is: Isomap
QUESTION 19
Pasting and bagging can be done on same algorithm.
a.
False
b.
True
Feedback
The correct answer is: True
QUESTION 20
Distance between the support vectors is called:
a.
Hyperplane
b.
Margin
c.
Response
d.
Separator
Feedback
The correct answer is: Margin
QUESTION 21
a.
False
b.
True
c.
Depends on Data
d.
Depends on Model
Feedback
The correct answer is: True
QUESTION 22
The error rate on old cases is called............
a.
Normalization error
b.
Overfitting error
c.
Generalization error
d.
None of these
Feedback
The correct answer is: None of these
QUESTION 23
Variables in Python are case-sensitive name.
a.
False
b.
Depends on Editor
c.
True
Feedback
The correct answer is: True
QUESTION 24
A multinomial classifier can be used as binomial classier.
a.
Depends on Data
b.
Always False
c.
Only if Data is Multi-label
d.
Always True
Feedback
The correct answer is: Always True
QUESTION 25
a.
Import modules
b.
Import workspace
c.
All of these
d.
Import libraries
Feedback
The correct answer is: All of these
QUESTION 26

>>> X, y = mnist["data"], mnist["target"] >>>, X.shape(70000, 784). Therefore, choose


X.shape(?, ?)
a.
X.shape(images, feature)
b.
Both a & b
c.
X.shape(feature, images)
d.
X.shape(MNIST, shape)
Feedback
The correct answer is: X.shape(images, feature)
QUESTION 27
Ensembles will yield bad results when there is significant diversity among the models
when all individual models have meaningful and good predictions.
a.
True
b.
False
c.
May or may not
Feedback
The correct answer is: False
QUESTION 28
A dense NumPy array would be our output when we are using.
a.
Both A & b
b.
LabelBinarize
c.
One-hot encoding
Feedback
The correct answer is: One-hot encoding
QUESTION 29
Binary classifiers cannot be combined to generate a multinomial classier.
a.
False
b.
Sometimes
c.
True
d.
Depends on Data
Feedback
The correct answer is: False
QUESTION 30
While creating workspace directory of our Machine Learning code and datasets. The
command is written.............. $ prompts.
a.
Before the
b.
After the
c.
At start of
d.
Both A & B
Feedback
The correct answer is: After the
QUESTION 31
Locally Linear Embedding (LLE) is a very powerful nonlinear dimensionality reduction
technique
a.
False
b.
True
Feedback
The correct answer is: True
QUESTION 32
Which of these statement(s) is/are true about k-NN algorithm?. (1) k-NN performs much
better if all of the data have the same scale. (2) k-NN works well with a small number of
input variables (p), but struggles when the number of inputs is very large. (3) k-NN
makes no assumptions about the functional form of the problem being solved
a.
1 and 3
b.
All 1, 2 and 3
c.
Only 1
d.
1 and 2
Feedback
The correct answer is: All 1, 2 and 3
QUESTION 33
Support vectors are the data points that lie closest to the decision surface.
a.
FALSE
b.
TRUE
Feedback
The correct answer is: TRUE
QUESTION 34
A hyper parameter is a parameter of learning model
a.
True
b.
False
Feedback
The correct answer is: False

Quiz 1 MCQ’s
QUESTION 1
Which one is true about neural networks:
a.
In artificial systems, neurons are not connected.
b.
In biological systems, neurons are not connected.
c.
In biological systems, neurons are connected much less than in artificial system.
d.
In biological systems, neurons are connected much more than in artificial system.
Feedback
The correct answer is: In biological systems, neurons are connected much more than in
artificial system.
QUESTION 2
While XOR classification problem output will be 1 if the inputs are:
a.
(0,1)
b.
(0,0)
c.
Both B & C
d.
(1,0)
Feedback
The correct answer is: Both B & C
QUESTION 3
Artificial neuron ............... binary input and ................ binary output
a.
One or more, greater than 1
b.
One, more than 1
c.
one or more, one
d.
One or more, one or more
Feedback
The correct answer is: one or more, one
QUESTION 4
While logical computations with neurons, neuron C gets activated if either neuron A or
neuron B is activated, which logic will be used to do so?
a.
OR
b.
NAND
c.
NOT
d.
AND
Feedback
The correct answer is: OR
QUESTION 5
Which package is consider standard for storing numerical data.
a.
None
b.
Pandas
c.
NumPy\Pandas
d.
Numpy
Feedback
The correct answer is: Numpy
QUESTION 6
In perception algorithm penalty is denoted as:
a.
None
b.
eta0
c.
Constant
d.
eta1
Feedback
The correct answer is: eta0
QUESTION 7
Artificial Neural Networks are composed of a large number of highly interconnected
processing elements called
a.
Cells
b.
Connections
c.
Blocks
d.
Neurons
Feedback
The correct answer is: Neurons
QUESTION 8
In ANNs the information passing between neurons is
a.
Same as in the biological counterparts
b.
Slow as compare to the biological counterparts
c.
None of the mentioned
d.
Fast as compare to the biological counterparts
Feedback
The correct answer is: Fast as compare to the biological counterparts
QUESTION 9
Feedforward neural network the signal is ........... directional
a.
Uni
b.
All of these
c.
Multi
d.
Arbitrary
Feedback
The correct answer is: Uni
QUESTION 10
Which key word is used to pandas in python.
a.
Pd
b.
None
c.
Pad
d.
pd
Feedback
The correct answer is: pd
QUESTION 11
If decision boundary is non-linear so the perceptions are ................... of learning
complex pattern
a.
Capable
b.
None of these
c.
Both A & B
d.
Incapable
Feedback
The correct answer is: Capable
QUESTION 12
Numpy is also consider essential for other packages
a.
Pandas\Scikit-learn
b.
Pandas
c.
Scikit-learn
d.
Pandas
Feedback
The correct answer is: Scikit-learn
QUESTION 13
Learning in biological systems involves
a.
Addition of new neurons and connections to the system
b.
None of the mentioned
c.
Adjustments to the synaptic connections that exist between the neurons
Feedback
The correct answer is: Adjustments to the synaptic connections that exist between the neurons
QUESTION 14
When a data scientist needs to deals with two dimentional data in python, than which
package is preferable
a.
None
b.
Numpy
c.
Scikit-learn
d.
Pandas
Feedback
The correct answer is: Pandas
QUESTION 15
An ANN can be characterized by:
a.
Architecture, Neurons, Layers
b.
Architecture, Neurons, Activation Function
c.
Architecture, Training or Learning Algorithms, Neurons
d.
Architecture, Training or Learning Algorithms, Activation Function
Feedback
The correct answer is: Architecture, Training or Learning Algorithms, Activation Function

Mohsin Ali
While logical computations with neurons, neuron C gets activated if either neuron A or neuron
B is activated, which logic will be used to do so?
a.OR
b.AND
c.NOT
d.NAND
Feedback
The correct answer is: OR

QUESTION 2
Artificial neuron ............... binary input and ................ binary output
a. One or more, greater than 1
b. One or more, one or more
c. One, more than 1
d. one or more, one
Feedback
The correct answer is: one or more, one
QUESTION 3
Numpy is also consider essential for other packages
a. Pandas
b. Pandas\Scikit-learn
c. Scikit-learn
d. Pandas
Feedback
The correct answer is: Scikit-learn

QUESTION 4
Which one is true about neural networks:
a. In artificial systems, neurons are not connected.
b. In biological systems, neurons are connected much more than in artificial system.
c. In biological systems, neurons are not connected.
d. In biological systems, neurons are connected much less than in artificial system.
Feedback
The correct answer is: In biological systems, neurons are connected much more than in
artificial system.
QUESTION 5
An ANN can be characterized by:
a. Architecture, Neurons, Activation Function
b. Architecture, Training or Learning Algorithms, Neurons
c. Architecture, Neurons, Layers
d. Architecture, Training or Learning Algorithms, Activation Function
Feedback
The correct answer is: Architecture, Training or Learning Algorithms, Activation Function
QUESTION 6
When a data scientist needs to deals with two dimentional data in python, than which package is
preferable
a. Pandas
b. Numpy
c. Scikit-learn
d. None
Feedback
The correct answer is: Pandas

QUESTION 7
In ANNs the information passing between neurons is
a. Slow as compare to the biological counterparts
b. Same as in the biological counterparts
c. Fast as compare to the biological counterparts
d. None of the mentioned
Feedback
The correct answer is: Fast as compare to the biological counterparts

QUESTION 8
Feedforward neural network the signal is ........... directional
a. All of these
b. Multi
c. Uni
d. Arbitrary
Feedback
The correct answer is: Uni
QUESTION 9
Which package is consider standard for storing numerical data.
a. Numpy
b. Pandas
c. NumPy\Pandas
d. None
Feedback
The correct answer is: Numpy
QUESTION 10
Artificial Neural Networks are composed of a large number of highly interconnected processing
elements called
a. Cells
b. Connections
c. Blocks
d. Neurons
Feedback
The correct answer is: Neurons
QUESTION 11
Which key word is used to pandas in python.
a. Pd
b. Pad
c. None
d. pd
Feedback
The correct answer is: pd
QUESTION 12
If decision boundary is non-linear so the perceptions are ................... of learning complex
pattern
a. Both A & B
b. Capable
c. None of these
d. Incapable
Feedback
The correct answer is: Capable
QUESTION 13
Learning in biological systems involves
a. Adjustments to the synaptic connections that exist between the neurons
b. None of the mentioned
c. Addition of new neurons and connections to the system
Feedback
The correct answer is: Adjustments to the synaptic connections that exist between the neurons
QUESTION 14
In perception algorithm penalty is denoted as:
a. eta1
b. Constant
c. None
d. eta0
Feedback
The correct answer is: eta0
QUESTION 15
Question text
While XOR classification problem output will be 1 if the inputs are:
a. (0,1)
b. (0,0)
c. (1,0)
d. Both B & C
Feedback
The correct answer is: Both B & C

Binary classifiers cannot be combined to generate a multinomial classier.


a. False
b. True
c. Depends on Data
d. Sometimes
Feedback
The correct answer is: False
QUESTION 2
A multinomial classifier can be used as binomial classier.
a. Always False
b. Only if Data is Multi-label
c. Depends on Data
d. Always True
Feedback
The correct answer is: Always True
QUESTION 3
Ensembles will yield bad results when there is significant diversity among the models when all
individual models have meaningful and good predictions.
a. False
b. True
c. May or may not
Feedback
The correct answer is: False
QUESTION 4
Pasting and bagging can be done on same algorithm.
a. True
b. False
Feedback
The correct answer is: True
QUESTION 5
a. Creating new features
b. Feature extraction
c. All of these
d. Feature selection
Feedback
The correct answer is: All of these
QUESTION 6
a. 0 to 3
b. 1 to 3
c. 0 to 4
d. 1 to 4
Feedback
The correct answer is: 0 to 4
QUESTION 7
The error rate on old cases is called............
a. Normalization error
b. Overfitting error
c. None of these
d. Generalization error
Feedback
The correct answer is: None of these
QUESTION 8
Distance between the support vectors is called:
a. Separator
b. Margin
c. Hyperplane
d. Response
Feedback
The correct answer is: Margin
QUESTION 9
Classes that have same application programming interface are.
a. Extra trees regressor and random trees regressor
b. Description and detail
c. numpy.ndarray and iris.target
d. Sklearn.utils.Bunch and sklearn.utils.close
QUESTION 10
In an election, N candidates are competing against each other and people are voting for either of
the candidates. Voters don’t communicate with each other while casting their votes. => Now
which of the following ensemble method works similar to above-discussed election procedure?
Hint: Persons are like base models of ensemble method.
a. Boosting
b. Bagging
c. AdaBoost
d. Learn++
Feedback
The correct answer is: Bagging
QUESTION 11
The best dimensionality reduction algorithm is called
a. Isomap
b. PCA
c. LLE
d. LDA
Feedback
The correct answer is: Isomap
QUESTION 12
A dense NumPy array would be our output when we are using.
a. Both A & b
b. One-hot encoding
c. LabelBinarize
Feedback
The correct answer is: One-hot encoding
QUESTION 13
What is true about an ensembled classifier? (1) Classifiers that are more “sure” can vote with
more conviction. (2) Classifiers can be more “sure” about a particular part of the space. (3) Most
of the times, it performs better than a single classifier.
a. 2 and 3
b. 1 and 3
c. 1 and 2
d. All of the above
Feedback
The correct answer is: All of the above
QUESTION 14
Some telecommunication company wants to segment their customers into distinct groups in
order to send appropriate subscription offers, this is an example of
a. Unsupervised learning
b. Serration
c. Data extraction
d. Supervised learning
Feedback
The correct answer is: Unsupervised learning
QUESTION 15
Locally Linear Embedding (LLE) is a very powerful nonlinear dimensionality reduction
technique
a. False
b. True
Feedback
The correct answer is: True
QUESTION 16
When a classifier is trained, it stores the list of target classes in its classes_ attribute.
a. In descending order
b. Randomly
c. Ordered by value
d. In ascending order
Feedback
The correct answer is: Ordered by value
QUESTION 17
How can SVM be classified?
a. It is a model trained using supervised learning. It can be used for classification and regression.
b. It is a model trained using unsupervised learning. It can be used for classification and regression.
c. It is a model trained using unsupervised learning. It can be used for classification but not for
regression.
d. It is a model trained using unsupervised learning. It can be used for classification but not for
regression.
Feedback
The correct answer is: It is a model trained using supervised learning. It can be used for
classification and regression.
QUESTION 18
While creating workspace directory of our Machine Learning code and datasets. The command
is written.............. $ prompts.
a. At start of
b. After the
c. Both A & B
d. Before the
Feedback
The correct answer is: After the
QUESTION 19
a. Import modules
b. Import workspace
c. All of these
d. Import libraries
Feedback
The correct answer is: All of these
QUESTION 20
If I am using all features of my dataset and I achieve 100% accuracy on my training set, but
~70% on validation set, what should I look out for?
a. Overfitting
b. Nothing, the model is perfect
c. Data Partition
d. Underfitting
Feedback
The correct answer is: Overfitting
QUESTION 21
In precision and recall VS the threshold, precision sometimes go up when you raise the
threshold
a. False
b. True
Feedback
The correct answer is: False
QUESTION 22
Confusion Matrix returns
a. Dataset evaluators
b. Predictions
c. Both A & B
d. Confused values
Feedback
The correct answer is: Predictions
QUESTION 23
Select the commonly best ratio for training the data and holding the data, respectively.
a. 60% and 40%
b. 40% and 60%
c. 80% and 20%
d. 70% and 30%
Feedback
The correct answer is: 80% and 20%
QUESTION 24
Which of these statement(s) is/are true about k-NN algorithm?. (1) k-NN performs much better
if all of the data have the same scale. (2) k-NN works well with a small number of input
variables (p), but struggles when the number of inputs is very large. (3) k-NN makes no
assumptions about the functional form of the problem being solved
a. 1 and 2
b. 1 and 3
c. All 1, 2 and 3
d. Only 1
Feedback
The correct answer is: All 1, 2 and 3
QUESTION 25
Support vectors are the data points that lie closest to the decision surface.
a. FALSE
b. TRUE
Feedback
The correct answer is: TRUE
QUESTION 26
By considering a thumb rule, precision/recall is preferred whenever the ............ class is rare
a. Extrinsic
b. Positive
c. Negative
d. Intrinsic
Feedback
The correct answer is: Positive
QUESTION 27
While creating workspace directory of our Machine Learning code and datasets. The command
is written after the $ prompts
a. False
b. True
Feedback
The correct answer is: True
QUESTION 28
a. Depends on Model
b. True
c. Depends on Data
d. False
Feedback
The correct answer is: True
QUESTION 29
Choose the odd one.
a. Principal Component Analysis
b. Multidimensional Scaling
c. Support vector machine
d. Linear Discriminant Analysis
Feedback
The correct answer is: Support vector machine
QUESTION 30
Variables in Python are case-sensitive name.
a. False
b. Depends on Editor
c. True
Feedback
The correct answer is: True
QUESTION 31
By considering large data which one is preferable?
a. RandomizedSearchCV
b. SearchSpaceCV
c. EnsembleSearchCV
d. GridSearchCV
Feedback
The correct answer is: RandomizedSearchCV
QUESTION 32
a. Both a & b
b. Transformer
c. Estimator
d. Predictor
Feedback
The correct answer is: Estimator
QUESTION 33
Train a binary classifier, capable of distinguishing between just two classes 6 and not-6. Choose
a random state.
a. y_train_5 = (y_train == 6) <= y_test_5 = (y_test == 6)
b. y_train_5 = (y_train == 6); y_test_5 = (y_test == 6)
c. y_train_6 = (y_train == 6) && y_test_6 = (y_test != 6)
d. y_train_5 = (y_train == 6) || y_test_5 = (y_test != 6)
Feedback
The correct answer is: y_train_5 = (y_train == 6); y_test_5 = (y_test == 6)
QUESTION 34
>>> X, y = mnist["data"], mnist["target"] >>>, X.shape(70000, 784). Therefore, choose
X.shape(?, ?)
a. X.shape(images, feature)
b. Both a & b
c. X.shape(MNIST, shape)
d. X.shape(feature, images)
Feedback
The correct answer is: X.shape(images, feature)

You might also like