12 Deep Learning Interview questions you should not be missed (Part 2)

5. How do you understand Backpropagation? Explain the mechanism of action?

This question aims to test knowledge of how a neural network works. You need to clarify the following points

• The Forward process (Forward Calculation) is a process that helps the model calculate the weights of each layer, and the resulting calculation will yield a yp result. This time will calculate the value of the loss function; the value of the loss function will show how good the model. If the loss function is not good enough, we need to find a way to reduce the value of the loss function. Training a neural network is substantially minimizing a loss function. The loss function L (yp, yt) represents the degree of difference between the output value of the yp model and the actual value of the yt data label.
• To reduce the value of the loss function, we need to use the derivative. Back-propagation helps us calculate the derivative for each layer of the network. Based on the value of the derivative on each layer, the optimizer (Adam, SGD, AdaDelta …) applying gradient descent will update the weight of the network.
• Back Propagation uses chain-rule mechanism or derivative function to calculate the gradient values of each layer…