Deep Learning

1.       What is the method to overcome the Decay of Information through time in RNN known as?

·          Gating

2.       The measure of Difference between two probability distributions is know as  

·         KL Divergence

3.       Recurrent Network can input Sequence of Data Points and Produce a Sequence of Output

·         True

4.       Prediction Accuracy of a Neural Network depends on _______________ and ______________

·          Weight & Bias

5.       Process of improving the accuracy of a Neural Network is called –

·         Training

6.       The rate at which cost changes with respect to weight or bias is called __________________

·         Gradient

7.       A _______________ matches or surpasses the output of an individual neuron to a visual stimuli. –

·         Convolution

8.       Autoencoders cannot be used for Dimensionality Reduction.

·         FALSE

9.       Data Collected from Survey results is an example of _ -

·         Structured Data

10.   Recurrent Networks work best for Speech Recognition. –

·         True

11.   How do RNTS interpret words?

·          Vector representation

12.   De-noising and Contractive are examples of

·         Autoencoders

13.   Gradient at a given layer is the product of all gradients at the previous layers.

·         True

14.   A Deep Belief Network is a stack of Restricted Boltzmann Machines.

·         True

15.   What are the two layers of a Restricted Boltzmann Machine called ?

·         Visible and Hidden

16.   All the Visible Layers in a Restricted Boltzmann Machine are connected to each other.

·         False

17.   Name the component of a Neural Network where the true value of the input is not observed.

·         Hidden Layer

18.   What is the difference between the actual output and generated output known as?

·         Cost

19.   In a Neural Network, all the edges and nodes have the same Weight and Bias values.

·         False

20.   What does LSTM stand for?

·         Long Short Term memory

21.   RELU stands for _______________

·         Rectified Linear Unit

22.   A Shallow Neural Network has only one hidden layer between Input and Output layers.

·         True

23.   Restricted Boltzmann Machine expect the data to be labeled for Training.

·         False

24.   Support Vector Machines, Naive Bayes and Logistic Regression are used for solving _______________________ problems.

·         Classification

25.   _____________ is a recommended Model for Pattern Recognition in Unlabeled Data.

·         Autoencoders

26.   Recurrent Neural Networks are best suited for Text Processing.

·         True