Skip to content

FRM Part I · FRM Exam Part I · Machine Learning and Prediction

A neural network with many parameters achieves very low error on the training set but substantially higher error on a validation set. Which action is most appropriate to address this problem?

Apply regularization such as dropout or early stopping. Low training error with high validation error signals overfitting, and these techniques limit model complexity or halt training before noise is fitted. Adding layers or training longer would worsen it, and dropping validation data would conceal it.

  1. AApply regularization such as dropout or early stoppingCorrect
  2. BAdd more hidden layers and nodes
  3. CTrain for more epochs without monitoring validation error
  4. DRemove the validation set and rely on training error

Explanation

The pattern indicates overfitting: the model fits noise in the training data. Regularization techniques such as dropout, weight penalties, or early stopping reduce complexity or stop training when validation error begins to rise. Adding capacity or training longer would worsen overfitting, and removing validation data hides the problem.

Did you get it right without looking?

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

Practise Machine Learning and PredictionFree diagnostic mock

More Machine Learning and Prediction questions