ARTIFICIAL INTELLIGENCE IMPORTANT TOPICS

What is Artificial Intelligence? Give an example of where AI is used

on a daily basis.

“Artificial Intelligence (AI) is an area of computer science that emphasizes the creation
of intelligent machines that work and react like humans.” “The capability of a
machine to imitate the intelligent human behavior.”
What are the different types of AI?

● Reactive Machines AI: Based on present actions, it cannot use previous

experiences to form current decisions and simultaneously update their memory.
Example: Deep Blue
● Limited Memory AI: Used in self-driving cars. They detect the movement of
vehicles around them constantly and add it to their memory.
● Theory of Mind AI: Advanced AI that has the ability to understand emotions,
people and other things in the real world.
● Self Aware AI: AIs that posses human-like consciousness and reactions. Such
machines have the ability to form self-driven actions.
● Artificial Narrow Intelligence (ANI): General purpose AI, used in building virtual
assistants like Siri.
● Artificial General Intelligence (AGI): Also known as strong AI. An example is the
Pillo robot that answers questions related to health.
● Artificial Superhuman Intelligence (ASI): AI that possesses the ability to do
everything that a human can do and more. An example is the Alpha 2 which is the
first humanoid ASI robot.

How is Machine Learning related to Artificial Intelligence?

Artificial Intelligence is a technique that enables machines to mimic human behavior.

Whereas, Machine Learning is a subset of Artificial Intelligence. It is the science of
getting computers to act by feeding them data and letting them learn a few tricks on
their own, without being explicitly programmed to do so.

What is Q-Learning?

The Q-learning is a Reinforcement Learning algorithm in which an agent tries to

learn the optimal policy from its past experiences with the environment. The past
experiences of an agent are a sequence of state-action-rewards:
 The probability density function and probability mass function

Continuous variable

Discrete variable

Probability distributions for Continuous variables

Probability distributions

which Fuzzy logic is an extension to the Crisp set, which handles the Partial Truth.

True

False

A 3-input neuron is trained to output a zero when the input is 110 and a one when the
input is 111. After generalization, the output will be zero when and only when the input
is?

A. 000 or 110 or 011 or 101

B. 010 or 100 or 110 or 101

C. 000 or 010 or 110 or 100

D. 100 or 111 or 101 or 001

Answer» C. 000 or 010 or 110 or 100

What is an auto-associative network?

a neural network that contains no loops
b neural network that contains feedback
c neural network that has only one loop
d single layer feed-forward neural network with pre-processing
 A 3-input neuron has weights 1, 2, 3. The transfer function is linear with the constant of
proportionality being equal to 3. The inputs are 4, 20, 5 respectively. What will be the
output?
ans→177

In Feedforward ANN, define the information flow

A. unidirectional
B. bidirectional
C. multidirectional
D. All of the above
What is the abbreviation of JPEG?
ans→Joint Photographic Experts Group

What are the Applications of Neural Networks?

A. sales forecasting

B. data validation

C. risk management

D. all of the mentioned

The network that involves backward links from output to the input and hidden layers
a) Self organizing maps
b) Perceptrons
c) Recurrent neural network
d) Multi layered perceptron
ans C

What is Image enhancement? What are its process

the procedure of improving the quality and information content of original
data before processing

What is Deep Learning?

Deep learning imitates the way our brain works i.e. it learns from experiences. It uses
the concepts of neural networks to solve complex problems.

​ Which characteristics are taken together in chromaticity? What is chromaticity

​ a) Saturation and Brightness
​ b) Hue and Saturation
​ c) Hue and Brightness
​ d) Saturation, Hue and Brightness
ans B

​ Which of the following is the next step in image processing after compression?
a) Wavelets
b) Segmentation
c) Representation and description
d) Morphological processing

ans D
​ What is Structuring element
​ (a)pixels
​ (b)lines
​ (c)subimage
​ (d)noise
​ ans C
​

​ Dilation process is used for what process in images

​

a.
thinner

shrinked
b.

thickened
c.

sharpened
d.

Answer:thickened

​ At what points, a continuous image is digitized?

​ a) random
​ b) vertex
​ c) contour
​ d) sampling
​ ans D

​ Which of the following correctly describes the slightest visible change in the level
 of intensity?
​ Contour

​ Saturation

​ Contrast

​ Intensity Resolution

​ ans intensity resolution

​ What is the name of the tool that helps in zooming, shrinking, rotating, etc.?
​ Filters

​ Interpolation

​ Sampling

​ None of the above

​ ans Interpolation
​
​ Explain the commonly used Artificial Neural Networks.

Feedforward Neural Network

​ The simplest form of ANN, where the data or the input travels in one direction.

​ The data passes through the input nodes and exit on the output nodes. This neural
network may or may not have the hidden layers.

Convolutional Neural Network

​ Here, input features are taken in batch wise like a filter. This will help the network to
remember the images in parts and can compute the operations.
​ Mainly used for signal and image processing

Recurrent Neural Network(RNN) – Long Short Term Memory

​ Works on the principle of saving the output of a layer and feeding this back to the
input to help in predicting the outcome of the layer.
​ Here, you let the neural network to work on the front propagation and remember
what information it needs for later use
​ This way each neuron will remember some information it had in the previous time-
step.

Autoencoders
​ These are unsupervised learning models with an input layer, an output layer and one
or more hidden layers connecting them.
​ The output layer has the same number of units as the input layer. Its purpose is to
reconstruct its own inputs.
​ Typically for the purpose of dimensionality reduction and for learning generative
models of data.

​ Define The dynamic range of the imaging system is a quantitative relation where
the upper limit can be determined by
​ Brightness

​ Contrast

​ Saturation

​ Noise

​ ans C
​ What is Image Multiplication?

​ What is the relationship between wavelength and frequency?

​ A. C = v + λ
​ B. C = λ/v
​ C. C = vλ
​ D. C = 1/λv
​ Answer: Option C
​
​ How to carry out an array function together with one or more images?
​ a) array by array
​ b) pixel by pixel
​ c) column by column
​ d) row by row
​ ans B
​ What is meant by Region of Interest (ROI) operations?
​ a) Shading correction
​ b) Masking
​ c) Dilation
​ d) None of the Mentioned
​ ans B
Blurring an image with the help of a smoothing filter may lead to noise reduction.
 True

False

ans A
​ application of image blurring?
​ Gross representation

​ Object motion

​ Object detection

​ Image segmentation

​ ans A
​ How to resolve the dark features in the image?
​ Gaussian Transform

​ Laplacian Transform

​ Power-law Transformation

​ Histogram Specification

​ ans C
​ What is gradient image ?
​ An image gradient is a directional change in the intensity or color in an image.

​ what is the smallest possible value of the gradient image?

​ 0
​ What is the name of the process that moves a filter mask over the image, followed by
calculating the sum of products?
ans correlation

Which of the following fact is true for an image?

a) An image is the addition of illumination and reflectance component

b) An image is the subtraction of illumination component from reflectance component
c) An image is the subtraction of reflectance component from illumination component
d) An image is the multiplication of illumination and reflectance component
ans D

Which of the following grey level intensities help in increasing the grey levels dynamic range
in the image?

a. Contrast Stretching

b. Negative Transformations
 c. Power-law Transformations

d. None of the above

ans A

open-source machine learning library for Python and is completely based on Torch

A. Flask
B. PyTorch
C. PyBrain
D. CherryPy
ans B

Which of the following are the advantages of PyTorch?

A. It is easy to debug and understand the code
B. It includes many layers as Torch
C. It includes lot of loss functions
D. All of the above
ans D

an imperative n-dimensional array which runs on GPU.

a. Variable
b. Tensor
c. Module
d. All of the above

a package which is used to wrap a tensor.

Variable

a technique or way to find the linear relation between the dependent variable and the
independent variable by minimizing the distance.

⮚ MSE stands for?

A. Machine Squared Error
B. Mean Simple Error
C. Mean Squared Evaluate
D. Mean Squared Error
ans D

⮚ The activation functions are basically divided

A. 1
B. 2
C. 3
D. 4
 ans 2

⮚ The Conv1d and Conv2D is used to apply 1D and 2D convolution.

A. TRUE
B. FALSE
C. Can be true or false
D. Can not say
ans A
⮚ The torch.cuda is a package which adds support for CUDA tensor type.
A. TRUE
B. FALSE
C. Can be true or false
D. Can not say
ans A

⮚ Which of the following is a subset of machine learning?

A. Numpy
B. SciPy
C. Deep Learning
D. All of the above
ans C

⮚ How many layers Deep learning algorithms are constructed?

A. 2
B. 3
C. 4
D. 5
ans B

⮚ The first layer is called the?

A. inner layer
B. outer layer
C. hidden layer
D. None of the above
ans A

⮚ RNNs stands for?

A. Receives neural networks
B. Report neural networks
C. Recording neural networks
D. Recurrent neural networks
ans D

⮚ Which of the following is/are Common uses of RNNs?

A. BusinessesHelp securities traders to generate analytic reports
B. Detect fraudulent credit-card transaction
 C. Provide a caption for images
D. All of the above
ans D
⮚ Which of the following is well suited for perceptual tasks?
A. Feed-forward neural networks
B. Recurrent neural networks
C. Convolutional neural networks
D. Reinforcement Learning
ans C
➢ CNN is mostly used when there is an?
A. structured data
B. unstructured data
C. Both A and B
D. None of the above
ans B

⮚ Which neural network has only one hidden layer between the input and output?
A. Shallow neural network
B. Deep neural network
C. Feed-forward neural networks
D. Recurrent neural networks
ans A
⮚ Which of the following is/are Limitations of deep learning?
A. Data labeling
B. Obtain huge training datasets
C. Both A and B
D. None of the above
ans C
⮚ Deep learning algorithms are more accurate than machine learning algorithm in image
classification.
A. 33%
B. 37%
C. 40%
D. 41%
ans D

⮚ Back propagation is a learning technique that adjusts weights in the neural network by
propagating weight changes.

a.
Forward from source to sink

b Backward from sink to source

.
 c Forward from source to hidden nodes
.

d Backward from sink to hidden nodes

.

⮚ Which of the following neural networks uses supervised learning?

(A) Multilayer perceptron

(B) Self organizing feature map

(C) Hopfield network

⮚ What are the tasks that cannot be realised or recognised by simple networks?
A. handwritten characters
B. speech sequences
C. image sequences
D. all of the mentioned

⮚ Activation value is associated with?

A. potential at synapses
B. cell membrane potential
C. all of the mentioned
D. none of the mentioned
⮚ What are 3 basic types of neural nets that form basic functional units among
i)feedforward ii) loop iii) recurrent iv) feedback v) combination of feed forward
& back
a) i, ii, iii
b) i, ii, iv
c) i, iv, v
d) i, iii, v
⮚ Feedback networks are used for auto association & pattern storage?
(a) yes
(b) no

⮚ Who invented perceptron neural networks?

 Frank Rosenblatt
⮚ What was the 2nd stage in perceptron model called?
a) sensory units
b) summing unit
c) association unit
d) output unit
 Consider the following statements regarding Artificial Neural Networks (ANN) and
Convolutional Neural Networks (CNN)
1.There are sparse connections between inputs and outputs between two
consecutivelayers in a CNN.
2.CNNs can be used only for image data
3.Parameters are shared between output neurons in a CNN layer.
4.Both CNNs and ANNs can take image data as input

There are sparse connections between inputs and outputs between two consecutive layers
in a CNN
Parameters are shared between output neurons in a CNN layer.
There are sparse connections between inputs and outputs between two consecutive layers
of an ANN
1. For any two layers with the same number of neurons an ANN will have fewer
parameters than a CNN

2. Which of the following is true for most CNN architectures?

A) It can help in dimensionality reduction

B) It can be used for feature pooling
C) It suffers less overfitting due to small kernel size
D) All of the above
3. Pooling layers are used to accomplish which of the following?
reduce the dimensions of the feature maps

In the context of deep learning, transfer learning depends on which of the following
variables:

Choose the correct statement with regard to different approaches used for hyper-parameter
optimisation
The ML engineer in the company decides that using the full, unprocessed video is too
expensive for an RNN. So, she decides to encode the image by using a CNN architecture
similar to Alexnet. The final, fully connected layer in her architecture has size 50. She uses
this embedding (final layer) as the input to the RNN for each frame. She also uses a single
hidden layer with 100 neurons. Answer the following questions.
1. What is the size of the bias matrix for the recurrent unit?
2. K-means is not deterministic and it also consists of number of iterations.
a) True
b) False
ans A
3. Which of the following can act as possible termination conditions in
 K-Means?
1. For a fixed number of iterations.

2. Assignment of observations to clusters does not change

between iterations. Except for cases with a bad local minimum.

3. Centroids do not change between successive iterations.

4. Terminate when RSS falls below a threshold.

A. 1, 3 and 4

B. 1, 2 and 3

C. 1, 2 and 4

D. 1,2,3,4

Answer» D. 1,2,3,4

4. Which of the following statements are true with regard to K-Nearest

Neighbours?

It provides excellent run-time performance.

✖

Its computation complexity is very low.

✔

It is easy to implement.
✔

It is robust to noisy training data.

ye bhi correct h 2,3,4

,
⮚ What are the objectives of exploratory data analysis?
>>All of the above.
 ⮚ What is Exploratory Data Analysis?
A rigid framework by which we analyze data
An initial way by which we can get a feel for data
A type of purely quantitative method of data analysis
A set of scientific principles for analyzing data in a categorical manner

⮚ What is used to graph can be used for simple summarization of data?

A. Scatterplot

B. Overlaying

C. Barplot

D. All of the Mentioned

⮚ What is the role of exploratory graphs in data analysis?

a) They are made for formal presentations
b) They are typically made very quickly
c) Axes, legends, and other details are clean and exactly detailed
d) They are used in place of formal modeling
⮚ What is used activation function is generally categorized into:

⮚ The network that involves backward links from output to the input and hidden layers is
called
a) Self organizing maps
b) Perceptrons
c) Recurrent neural network
d) Multi layered perceptron
ans C

⮚ What is used to calculate Loss function in PyTorch?

>>Hinge Embedding Loss Function.

⮚ The numbers of pooling layers in PyTorch are:

⮚ What does the import cv2 statement do?

>> Import our openCV python Bindings.

Imports the SciPy library for numerical processing.

Imports the NumPy library for numerical processing.

Imports our OpenCV Python bindings.

Displays an image to our screen.

⮚ Suppose our image has a width of 700 pixels, a height of 550 pixels, and 3 channels,
one for each Red, Green, and Blue component. How would we express this image as a
NumPy array shape?
(a) (550,700,3)
(b) (3,550,700)
(c) (700,550,3)
(d) (3,700,550)

⮚ What is RGB tuple codes for blue.

(0,0,255)

The RGB tuple (255, 0, 0) codes for red. But OpenCV

would actually interpret this color as:
ans blue
⮚ Suppose, We have an image whose resolution is 760x 512. How many total pixels are
in the image?
ans 389120

⮚ The process of converting an image from other colour space, varying the pixel
intensity in between the black and white region is known as:
Grayscaling

⮚ What is used to used to draw line in OpenCV?

cv2. line()

⮚ What is used to OpenCV is used to detect edges of the image?

The Sobel is a filter in the OpenCV to detect the edge of an image.
ans cv2.Canny()
⮚ What is used to OpenCV is used to show an output image?
cv2. imshow()

⮚ The average positive difference between computed and desired outcome values is
known as:
(a) Root mean squared error
(b) Mean squared error
(c) Mean absolute error
(d) Mean positive error
ans C
 ⮚ Which of the following statements is true for the Perceptron Learning Algorithm?

a. The Perceptron always converges to the best linear separator for a given dataset.

b. The convergence criteria for Perceptron depends on the initial value of the weight vector.

c. If the dataset is not linearly separable, the Perceptron algorithm does not converge and keeps cycling

between some sets of weights.

d. If the dataset is not lineary separable, the Perceptron algorithm learns the linear separator with

least misclassifications.

⮚ What is Sigmoid function

A. regression.

B. logistic.

C. probability.

D. neural.

Answer» B. logistic.

⮚ we use a learning rate that is too large in neural network? What will be the outcome
A. Network will converge

B. Network will not converge

C. Can’t Say

Solution: B
 ⮚ Consider the scenario. The problem you are trying to solve has a small amount of
data. Fortunately, you have a pre-trained neural network that was trained on a similar
problem. Which of the following methods would you choose to make use of this pre-
trained network?
A. Re-train the model for the new dataset

B. Assess on every layer how the model performs and only select a few of

them

C. Fine tune the last couple of layers only

D. Freeze all the layers except the last, re-train the last layer

Solution: (D)

⮚ Suppose that you are training a neural network for classification, but you have notice
that the training loss is much lower than the validation loss. How do u resolve it

⮚ What is used to representation of learning algorithm?

A) Neural network

B) Random Forest

C) k-Nearest neighbor

D) None of the above

Solution: (A)
What is single layer Perceptron?
a) a single layer feed-forward neural network with pre-processing
b) an auto-associative neural network
c) a double layer auto-associative neural network
d) a neural network that contains feedback
ans A