Prediction interval for neural network models using weighted asymmetric loss functions.
Abstract: We propose a simple and efficient approach to generate prediction intervals (PIs) for approximated and forecasted trends. Our method leverages a weighted asymmetric loss function to estimate the lower and upper bounds of the PIs, with the weights determined by the interval probability level. We provide a concise mathematical proof of the method, show how it can be extended to derive PIs for parametrised functions and argue why the method works for predicting PIs of dependent variables. The presented tests of the method on a real-world forecasting task using a neural network-based model show that it can produce reliable PIs in complex machine learning scenarios.
Submission Length: Regular submission (no more than 12 pages of main content)
Changes Since Last Submission: Changed a typo in the abstract.
Assigned Action Editor: ~Jasper_Snoek1
Submission Number: 1303
Loading