Forward propagation in ML Python - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is forward propagation in a neural network?

Forward propagation is the process where input data passes through the network layers to produce an output or prediction.

beginner

Why do we use activation functions during forward propagation?

Activation functions add non-linearity to the network, allowing it to learn complex patterns beyond simple linear relationships.

beginner

In forward propagation, what is the role of weights and biases?

Weights scale the input signals, and biases shift them. Together, they help the network adjust outputs to better fit the data.

intermediate

What happens if you skip forward propagation in training a neural network?

Without forward propagation, the network cannot produce predictions or calculate errors, so it cannot learn or improve.

beginner

How is the output of one layer used in forward propagation?

The output of one layer becomes the input to the next layer, passing information forward through the network.

What is the first step in forward propagation?

A Passing input data into the network

B Calculating the loss

C Updating weights

D Applying backpropagation

Forward propagation starts by feeding input data into the network to begin calculations.

Which of these is NOT part of forward propagation?

A Adjusting weights based on error

B Calculating weighted sums

C Applying activation functions

D Passing outputs to next layer

Adjusting weights happens during backpropagation, not forward propagation.

Why do we apply activation functions during forward propagation?

A To reduce input size

B To calculate loss

C To add non-linearity

D To update weights

Activation functions add non-linearity, enabling the network to learn complex patterns.

What does the output of the last layer represent after forward propagation?

A The input data

B The error value

C The updated weights

D The network's prediction

The last layer's output is the network's prediction based on the input.

Which components are combined to calculate the input to a neuron during forward propagation?

A Learning rate and momentum

B Weights, biases, and inputs

C Activation and error

D Loss and accuracy

Inputs are multiplied by weights and added to biases to calculate the neuron's input.

Explain the steps of forward propagation in a simple neural network.

Think about how data moves from input to output through each layer.

Why is forward propagation important for training a neural network?

Consider what would happen if the network could not produce outputs.