Quiz
Module 12 Review
Create a free account to finish this quiz
Save your progress and see your final score.
Already have an account? Sign in
Questions in this quiz (14)
1The Adam optimizer assigns weights to the parameters of the model in a __ way.
- Aadaptive
- Brecursive
- Coptimized
- Diterative
2The Adam Optimizer adds __ new hyperparameters to the model.
- Atwo
- Binfinite
- Czero
- Dthree
3In feed forward neural network, the output is compared with ___ before sending back
- APredicted value
- BActual value
- Cweights
- DOptimal weights
4Deep learning in contrast with machine learning will not need the following
- ADomain knowledge
- BLarge data set
- CCost function
- DClean data
5A feed forward neural network has its output values restricted to:
- A(0,1)
- B(-1,1)
- CNo restriction
- DPositive values only
6The MNIST dataset typically has these qualities:
Input layer: 784
Hidden layer: 100
Output layer: 10
How many total parameters are there?- A79,510
- B80,294
- C79,400
- D784,000
7Output from feed forward is compared with the actual value and the error is sent __________?
- ABackwards
- BForwards
- CSideways
- DUpwards
8Batch Normalization standardizes each piece of input data by _______ its mean and ________ by its standard deviation
- Asubtracting, dividing
- Badding, multiplying
- Csubtracting, multiplying
- Dadding, dividing
9Dropout regularization ________.
- Ais computationally inexpensive
- Bimpacts few parameters
- Cis applied for training and validation data
- Dremoves two or more hidden neurons during an epoch
10Aspects associated with deep learning models are _____, ____, and _____.
- Astopped training, LBFGS, normalized variance weight initialization
- Bdropout regularization, batch GD, BFGS
- CRectified linear activation function, constant variance weight initialization, Stochastic GD
- DGPU, hyperbolic tangent activation function, and batch normalization
11What is a dead unit in neural network?
- AA unit which doesn’t update during the training by any of its neighbour.
- BA unit which doesn’t respond to any of the training patterns.
- CThe unit which produces the biggest sum-squared error.
12Which of the following gives a non-linearity to a neural network?
- ARectified Linear Unit.
- BStochastic gradient Descent.
- CRectified Linear Unit
- DConvolution function
13What kind of relation does log-likelihood have with Chi-Square?
- AInverse
- BDirect
- CNone
14Which of the following in a CNN reduces each channel in the feature map to a single value?
- APooling
- BFully Connected Layer
- CTransfer Function
Information
- Description
- Quiz on Module 12 Review, created by Nick Feldman on 01/05/2022.
- Author
- Nick Feldman
- Language
- English
- Views
- 0
- Rating
- No ratings yet Rate this quiz