Quiz: How Neural Networks Learn (Forward & Backward Propagation)

Question 1 of 8

What is Forward Propagation?

The process where data flows from input layer through hidden layers to output layer

The process of adjusting weights after calculating error

The method of calculating the cost function

The technique of randomly initializing weights

Question 2 of 8

What formula is used to calculate the error in the cost function shown in the content?

Error = (Correct Answer - Prediction)²

Error = (Prediction - Correct Answer)²

Error = |Correct Answer - Prediction|

Error = (Correct Answer × Prediction)²

Question 3 of 8

What analogy can be used to explain Gradient Descent?

Being on a hill in the dark and taking small steps downward to reach the lowest point

Climbing a ladder step by step

Following a map to reach a destination

Swimming against a current

Question 4 of 8

What is the formula for updating weights during backward propagation? (Read through https://infolia.ai/archive/36 to answer)

New Weight = Old Weight - (Learning Rate × Gradient)

New Weight = Old Weight + (Learning Rate × Gradient)

New Weight = Old Weight × Learning Rate

New Weight = (Old Weight - Gradient) / Learning Rate

Question 5 of 8

In the weight update example provided, if the Old Weight is 0.5, Learning Rate is 0.1, and Gradient is 0.3, what is the New Weight? (Read through https://infolia.ai/archive/36 to answer)

0.47

0.53

0.50

0.80

Question 6 of 8

In which direction does Backward Propagation move through the network layers?

From output layer through hidden layers toward input layer

From input layer through hidden layers to output layer

Randomly between all layers

Only within the hidden layers

Question 7 of 8

What happens to the weights when a neural network first starts?

The weights are random

The weights are all set to zero

The weights are pre-trained and accurate

The weights are set to one

Question 8 of 8

What are the four steps that make up the complete learning process described in the content? (Read through https://infolia.ai/archive/36 to answer)

Forward Propagation, Measure Error, Backward Propagation, Repeat

Initialize Weights, Forward Propagation, Calculate Output, Stop

Load Data, Train Model, Test Model, Deploy

Input Data, Process Data, Output Results, Save Model

← Back to Newsletter