Quiz: Deep Learning and Neural Networks
Test your knowledge about deep learning and neural networks.
30 Questions - Click on each question to reveal the correct answer.
Question 1
What is Deep Learning?
A) Slow learning
B) A type of machine learning using deep neural networks (many layers)
C) Shallow learning
D) Learning without data
Show Answer
Correct Answer: B
Deep learning is a special type of machine learning that uses deep neural networks (many layers of neurons).
Question 2
What distinguishes deep learning from regular machine learning?
A) No difference
B) Deep learning uses many layers and discovers features by itself
C) Machine learning is always better
D) Deep learning is simpler
Show Answer
Correct Answer: B
Deep learning uses many layers and discovers important features by itself, while regular machine learning requires predefined features.
Question 3
What are neural networks?
A) Internet networks
B) A way to simulate how the human brain works using a computer
C) Telephone networks
D) Electrical networks
Show Answer
Correct Answer: B
Neural networks are a way to simulate how the human brain works using a computer - connected cells that process information.
Question 4
What are the basic components of neural networks?
A) Screen and keyboard
B) Neurons, connections, layers
C) Battery and charging
D) Camera and microphone
Show Answer
Correct Answer: B
The basic components are: Neurons, Connections, and Layers.
Question 5
What is the input layer in a neural network?
A) The layer that gives the result
B) The first layer that receives data
C) The middle layer
D) There is no such thing as an input layer
Show Answer
Correct Answer: B
The input layer is the first layer that receives raw data (like an image).
Question 6
What is the output layer in a neural network?
A) The first layer
B) The last layer that gives the result
C) The layer that receives data
D) There is no such thing as an output layer
Show Answer
Correct Answer: B
The output layer is the last layer that gives the final result (like "this is a cat").
Question 7
What are hidden layers?
A) Secret layers
B) Layers between input and output that process data
C) Non-existent layers
D) Layers hidden from the user
Show Answer
Correct Answer: B
Hidden layers are the layers between input and output that process data and discover patterns.
Question 8
How do neural networks learn?
A) By reading
B) By trying, comparing, correcting, and repeating
C) By listening
D) They don't learn
Show Answer
Correct Answer: B
Networks learn by: trying → comparing with the correct answer → correcting (adjusting weights) → repeating.
Question 9
What are "weights" in neural networks?
A) Computer weight
B) The strength of the connection between cells that controls information
C) Data weight
D) Nothing
Show Answer
Correct Answer: B
Weights are the strength of the connection between neurons - they control how information is passed and are adjusted during learning.
Question 10
In cat recognition, what does the first layer discover?
A) The whole cat
B) Lines and simple shapes
C) Colors only
D) Nothing
Show Answer
Correct Answer: B
The first layer discovers simple things: lines, curves, edges.
Question 11
What do the middle layers discover in cat recognition?
A) Simple lines only
B) Larger shapes like eyes and ears
C) The whole cat
D) Nothing
Show Answer
Correct Answer: B
The middle layers discover larger and more complex shapes like eyes, ears, and noses.
Question 12
What does the last layer do in cat recognition?
A) Detects lines
B) Decides "this is a cat" or "not a cat"
C) Collects data
D) Nothing
Show Answer
Correct Answer: B
The last layer gathers everything learned by previous layers and decides the final result.
Question 13
Why is the learning called "deep"?
A) Because it's difficult
B) Because it uses many (deep) layers
C) Because it's slow
D) Because it's old
Show Answer
Correct Answer: B
It's called "deep" because it uses deep neural networks - meaning many layers.
Question 14
What does deep learning need to work?
A) Only a little data
B) Lots of data, high computational power, time, expertise
C) Just a phone
D) Just the internet
Show Answer
Correct Answer: B
Deep learning needs: lots of data, high computational power, a long time for training, and expertise to build networks.
Question 15
Where do we find deep learning in phones?
A) Not in phones
B) Face recognition, voice assistants, translation
C) Charging only
D) Calling only
Show Answer
Correct Answer: B
In phones: face recognition, voice assistants (Siri, Google Assistant), instant translation.
Question 16
Why do neural networks mimic the human brain?
A) Because they look like it
B) Because they use connected cells to process information like the brain
C) Because they think
D) They don't mimic it
Show Answer
Correct Answer: B
They mimic the brain because they use interconnected cells, each receiving and sending information - like real neurons.
Question 17
What is the advantage of deep learning discovering features by itself?
A) No advantage
B) We don't need to define features beforehand, it discovers them
C) Makes it slower
D) Makes it weaker
Show Answer
Correct Answer: B
A big advantage: we don't need to tell the computer what the important features are, it discovers them by itself from the data.
Question 18
What is the "black box" in neural networks?
A) A real box
B) Difficulty in understanding exactly how the network makes decisions
C) Storage box
D) Nothing
Show Answer
Correct Answer: B
The "black box" means it's hard to understand exactly how the network makes decisions - the result is clear but the reason is obscure.
Question 19
How many layers are in a "deep" neural network?
A) One layer
B) Many layers (tens or hundreds)
C) Only two layers
D) No layers
Show Answer
Correct Answer: B
Deep networks contain many layers - from tens to hundreds of layers.
Question 20
Why is deep learning better at image recognition?
A) Because it's older
B) Because the many layers discover complex details
C) Because it's cheaper
D) Because it's simpler
Show Answer
Correct Answer: B
The many layers allow the network to discover complex and fine details in images.
Question 21
What are the challenges of deep learning?
A) No challenges
B) Needs lots of data, high cost, long time, complexity
C) Very easy
D) Doesn't work
Show Answer
Correct Answer: B
Challenges: needs lots of data, high cost for hardware, long training time, complex to understand.
Question 22
Does ChatGPT use deep learning?
A) No
B) Yes, it uses deep neural networks to understand and generate text
C) Uses regular programming only
D) Doesn't work
Show Answer
Correct Answer: B
Yes, ChatGPT uses deep neural networks (Transformers) to understand and generate text.
Question 23
What is the relationship between machine learning and deep learning?
A) No relationship
B) Deep learning is a type of machine learning
C) Machine learning is a type of deep learning
D) The same thing
Show Answer
Correct Answer: B
Deep learning is a special type of machine learning - a subset that uses deep neural networks.
Question 24
Why does deep learning need high computational power?
A) It doesn't
B) To process many layers and millions of computations
C) For decoration
D) For storage only
Show Answer
Correct Answer: B
It needs high computational power to process many layers and the millions or billions of required computations.
Question 25
What is the confidence level in neural networks?
A) Network weight
B) How sure the network is of its answer (like 95%)
C) Network speed
D) Data size
Show Answer
Correct Answer: B
Confidence level means how sure the network is of its answer - for example, "this is a cat with 95% confidence."
Question 26
Where is deep learning used in medicine?
A) Not used
B) Analyzing medical images, assisting in diagnosis, detecting diseases
C) Surgery only
D) Drugs only
Show Answer
Correct Answer: B
In medicine: analyzing medical images (X-rays), assisting in diagnosis, early disease detection.
Question 27
What is the difference between a simple and a deep neural network?
A) No difference
B) Deep networks have more layers and solve harder problems
C) Simple is better
D) Deep is older
Show Answer
Correct Answer: B
Deep networks have more layers and can solve more complex problems, but they require more resources.
Question 28
How does a neural network improve over time?
A) It doesn't improve
B) Adjusts weights based on errors to give better results
C) Buys an upgrade
D) Waits
Show Answer
Correct Answer: B
It improves by adjusting weights (connection strengths) based on errors, resulting in better outcomes.
Question 29
Modern machine translation uses:
A) Dictionaries only
B) Deep learning to understand context and translate naturally
C) Doesn't work
D) Word-for-word translation only
Show Answer
Correct Answer: B
Modern machine translation uses deep learning to understand context and translate naturally and understandably.
Question 30
What is the most important thing to remember about deep learning?
A) It's simple
B) It uses many layers to discover complex patterns and solve difficult problems
C) It doesn't work
D) It's old
Show Answer
Correct Answer: B
The most important thing: deep learning uses many layers of neurons to discover complex patterns and solve difficult problems.
Result
| Points | Level |
|---|---|
| 27-30 | Excellent - Strong understanding of deep learning |
| 21-26 | Good - Some review is helpful |
| 15-20 | Average - Review the chapters again |
| Less than 15 | Needs comprehensive review |
Concept Summary
| Concept | Explanation |
|---|---|
| Deep Learning | Machine learning with deep neural networks |
| Neural Networks | Connected cells simulating the brain |
| Layers | Input → Hidden → Output |
| Weights | Connection strength between cells |
| Training | Adjusting weights for improvement |