--------------------------------------------------------------------------------------------------------------------------
UNIT-1
1) What are Neural Networks? With a neat architecture diagram explain about Dense Deep Neural
Networks. (CO1, BL2)
2) What are Sparse Deep Neural Networks? Illustrate the working of sparse Deep Neural Networks with
the help of architecture diagram. (CO1, BL2)
3) Distinguish Sparse and Dense Neural Networks. (CO1, BL4)
4) Write the Advantages and Disadvantages of: (CO1, BL1)
a) Sparse Networks b) Dense Networks
5) Write the purpose of Activation functions. List and explain the working of different types of activation
functions with suitable examples. (CO1, BL2)
6) Tabulate the Differences of all activation functions. (CO1, BL3)
7) Specify the necessity of Hyperparameters in Deep learning. Outline the common examples of
Hyperparameters used in Deep learning. (CO1, BL3)
8) Define the term Hyperparameter Tuning? List and explain the different Hyperparameter Tuning
Techniques. (CO1, BL2)
9) Analyze and tabulate the differences of following Hyperparameter tuning Techniques: (CO1, BL3)
a) Manual Search b) Grid search c) Random search d) Bayesian Optimization
10) List the advantages and disadvantages of Hyperparameter Tuning Technique.
(CO1, BL1)
--------------------------------------------------------------------------------------------------------------------------
UNIT-II
1) What are Convolutional Neural Networks. List and explain about 1D and 2D Convolution
Dimensional Networks with suitable examples. (CO2, BL2)
2) Write the purpose of using conv2D in CNNs. With a neat Diagram Illustrate the working of conv2D.
(CO2, BL3)
3) What does layer +activation combination refer to in Deep learning. Discuss in detail about layer +
activation combination with an example. (CO2, BL2)
4) Outline the advantages of layer + activation combination in Deep Learning. (CO2, BL2)
5) Analyze and tabulate the differences of 1D, 2D CNNs. (CO2, BL3)
6) What does Pooling refer to in Deep Learning. List and explain the different types of Pooling strategies
used in Deep Learning with examples.
7) Explain in Detail about the sequence of operations performed in convolution layers from input image
to output feature map. (CO2, BL2)
8) Explain the following terms related to pooling strategies in Deep Learning: (CO2, BL2)
a) Max Pooling b) Average Pooling c) Stride
d) filter e) feature map f) Global Pooling
9) Explain in detail the concept of feature maps in CNNs. (CO2, BL2)
10) Given a 6×6 feature map, apply max pooling and average pooling using a 3×3 filter with stride 3. Show
intermediate and final outputs. (CO2, BL3)
UNIT-III
1) List the Key Characteristics of RNNs and explain the working of RNNs. (CO3, BL2)
2) What is LSTM. With a neat diagram explain the components of LSTM and also how LSTM Works.
(CO3, BL2)
3) Tabulate the differences of Traditional RNNs and LSTM. (CO3, BL4)
4) Write short notes about Sequence – to – vector models. (CO3, BL1)
5) Explain in detail about the use of Embedded layers in Deep Learning. (CO3, BL2)