Skip to content

FRM Part I · FRM Exam Part I · Machine-Learning Methods

A risk analyst builds a feedforward neural network to predict loan default. The network has an input layer, two hidden layers and an output layer. What is the primary role of the activation function applied at each hidden node?

The activation function introduces nonlinearity. Without it, multiple layers would collapse into one linear mapping, no better than linear regression. Nonlinear activations allow the network to capture complex interactions and patterns in the data, which is the reason hidden layers add modelling power.

  1. ATo introduce nonlinearity so the network can model complex relationshipsCorrect
  2. BTo reduce the number of input features before training begins
  3. CTo guarantee that the loss function has a single global minimum
  4. DTo scale the training labels to lie between zero and one

Explanation

Without a nonlinear activation, a stack of layers collapses into a single linear transformation, equivalent to linear regression. Activation functions let the network approximate nonlinear relationships. They do not reduce features, guarantee a convex loss, or rescale labels.

Did you get it right without looking?

One question tells you little. A timed set on Machine-Learning Methods shows your real accuracy, how long you take and where you lose marks.

Practise Machine-Learning MethodsFree diagnostic mock

More Machine-Learning Methods questions