Feedforward Network / Multi-layer Perceptron Quiz (Easy 1)

 What is the primary architectural characteristic of Multi-layer Perceptrons (MLPs) or a Feedforward Network (FNN)?

In the context of FNNs, what does the term “feed forward” mean?

In an FNN / MLP, what is the purpose of the “hidden layers”?

In an FNN / MLP, what is the purpose of the “output layer”?

What is the backpropagation algorithm used for in FNNs?

Which technique is commonly used to mitigate overfitting in MLPs?

In FNNs, what is the vanishing gradient problem, and how does it affect training?

How is the loss function typically defined in FNNs for regression tasks?

In MLPs, how is the loss function typically defined for binary classification tasks?

What is the purpose of the loss function in FNNs?

In FNNs, what is the purpose of the bias term associated with each neuron?