Q.1
14 Marks(a)
Define the term "information flow" in the context of neural networks.
(b)
Draw and explain basic architecture of ANN.
(c)
Explain the process of training an ANN, focusing on the role of the loss function and optimization algorithms.
Gujarat Technological UniversityWinter 2025 examination
GTU 116AG01 Deep Learning and Neural Network Winter 2025 Question Paper & PDF
B.E. · Computer Engineering · Semester 6 · Subject code 116AG01
- Total
- 70 marks
- To pass
- 23 marks (33%)
- Duration
- 2.5 Hours 2:30 PM – 5:00 PM
- Questions
- 5 with internal choice
No solutions on this paper yetNeed them for your exam? Ask once. Papers that students ask for are solved first, and we aim to add them within 24 hours.
Noted. This paper is in our queue now. Check back tomorrow.
We could not record that here. Email the paper name to hirenjsca@gmail.com and we will add it.
5 of 25 questions on this paper were also asked in other years. Each is marked below with the years it appeared.See all 14 repeated 116AG01 questions, most asked first
Q.2
14 Marks(a)
Explain the significance of weights sharing in RNN.
(b)
Explain the importance of multiple hidden layers in deep learning.
(c)
Describe the concept of "memory" in RNNs. Explain how it allows RNNs to process sequential data.
OR
(c)
Discuss the concept of time steps in RNNs. Explain How RNN processes sequential data differently than other neural networks.
Q.3
14 Marks(a)
Explain the role of resizing in image classification.
(b)
Outline the typical architecture of a CNN used for image classification.
Also asked in Summer 2026 Q3c, Summer 2025 Q4c, Summer 2024 Q3c OR
(c)
Explain the role of convolutional layers in a CNN. Discuss how convolutional filters work and the concept of feature maps.
OR
(a)
Explain the role of normalization in image classification.
(c)
Describe the purpose of pooling layers in CNNs. Compare max pooling and average pooling, including their effects on the network’s performance.
Q.4
14 Marks(a)
Give the difference between LSTM and RNN.
(b)
State applications of RNN in real world.
(c)
Discuss the role of each gate in an LSTM cell. Explain how each gate contributes to the flow of information.
OR
(a)
Compare LSTM with GRU.
(b)
Give limitations of LSTMs.
(c)
Provide an example of a scenario where LSTMs would be preferred over standard RNNs, and explain the reasoning.
Q.5
14 Marks(a)
Give advantages of TensorFlow.
(b)
Explain different layers in Keras.
OR
(a)
Give limitations of Keras.
More Deep Learning and Neural Network papers
Questions repeat across sessions. Working through two or three papers shows you which topics GTU asks most.
Other Semester 6 subjects
BE Computer
Preparing with friends?
Send this paper to your class group. Every question has its own link.
About this paper
Official Gujarat Technological University (GTU) examination paper for Deep Learning and Neural Network (Winter 2025, B.E. · Computer Engineering, Sem 6). Features complete 70-mark regular & remedial examination pattern, official marking distribution across all 5 questions, and direct 1-click official PDF download.
Transcribed for student exam preparation from Gujarat Technological University official examination archives. Questions, syllabus guidelines, and curriculum marking schemes remain the intellectual property of Gujarat Technological University.