CellStrat Artificial Intelligence Scholarship Test

JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

Email *

Which network is best suited for language translation ? *

CNN

SVM

LSTM

ANN

Transfer learning takes longer time than regular training ? *

True

False

TensorFlow-Slim offers *

Lighter version of TensorFlow

Pre-trained models

Both of the above

High-level TensorFlow APIs like Estimators and TF.Layers *

Make the code simpler

Make the computation faster

Make the code shorter

Both (a) and (c)

ANNs are defined as *

A technique for detecting fraud

A platform for Supervised Learning

A network of interconnected neurons

A network of biological objects

What is “deep” about deep learning ? *

Large number of training data

Large number of computational layers

Use of high powered CPU/GPU

None of the above

GAN uses two competing outputs *

True

False

Single Perceptron can create a curvilinear classifier *

True

False

Class scores are computed by *

Convolutional layer

Pooling layer

Fully connected layer

A scaled feature means that 10% change in feature of equal importance will produce equal change in output *

True

False

None of these

Mobile No *

This is a picture of *

Sigmoid function

KNN classifier

Logistic regression

Decision tree

This equation represents *

Median

Variance

Correlation

Standard deviation

x in the equation below is a *

scalar

vector

matrix

None of these

Which one of the following does not suffer from vanishing gradients as the input grows to infinity? *

Relu

Sigmoid

tanh

None of the above

Classification is used to find out *

Class of an object

Trend and the relations

Odd elements

Clusters

High learning rate means rapid gradient learning *

True

False

None of the Above

Overfitting leads to better generalization ? *

True

False

RNN can be a bi-directional network *

True

False

Picture on the right represents *

Low learning rate

High learning rate

Erratic learning rate

None of the above

The output of an activation function that takes a vector input is ? *

Scalar

Vector

Tensor

What RNN network is best suited for video captioning ? *

a) first one

b) second one

c) third one

d) fourth one

What is the correct order to solve a Machine learning Problem? *

Analysing the statement -> reviewing the error -> training -> testing

Analysing the statement -> training -> testing -> reviewing the error

Analysing the statement -> training -> reviewing the error -> testing

Analysing the statement -> testing -> training-> reviewing the error

Is the following depiction right? *

True

False

In the diagram on right, which lines represents the maximum margin hyperplane? *

The line cutting across with label “w”

The lines which contain red dots or green dots

The dotted line in the centre

None of the above

Why is a vanishing or exploding gradient a problem? *

No convergence

More memory requirement

Getting stuck in local minima

What is sigma in this equation: *

Variance

Standard Deviation

Mean

Co-Variance

TensorFlow Variable is typically used to *

Hold X and Y data sets

Hold weights and bias variables

Either of the above

None of the above

Which algorithm takes lowest amount of training time *

KNN

Logistic Regression

Decision Trees

Neural Networks

RNN is always a single layer network *

True

False

Which operator has higher precedence while executing a program *

% Modulus

& BitWise AND

** Exponent

> Comparison

Your Name *

When should we stop training the model ? *

Point A

Point B

Point C

Never

def calculate (num1, num2=4): *

error

What is Unsupervised Learning *

It is when humans are not available to code the algorithm

It is a method to learn patterns from data without knowing the ground-truth labels

It is a method to learn patterns from data when the ground-truth labels are known

It is a kind of learning which involves input and outputs of varied kinds

For an event E, P(E) + P (NOT E) = q, then *

0 ≤ q < 1

0 < q ≤ 1

0 < q < 1

None of these

In the evaluation phase which of the following is used? *

Forward propagation

Back propagation

Both of the above

None of the above

TensorFlow creates a computational graph *

True

False

Picture on the left represents *

Low learning rate

High learning rate

Erratic learning rate

None of the above

Determine the output of the following programming question *

WELCOME

welcome

Error

Infinite loop

Convolutional layer does the following *

Connects to the neurons of the next layers based on neighborhood principle

Connects with all the neurons in the next layer

Drops some weights

What is curse of dimensionality? *

It is a pervasive problem in real-life datasets

It is the problem of too many dimensions

It is the problem of too many manifolds

It is the problem of non-representative data

Submit

Clear form

Never submit passwords through Google Forms.

This content is neither created nor endorsed by Google. Report Abuse - Terms of Service - Privacy Policy

Forms