Remember to Correct the Bias When Using Deep Learning for Regression!

Research output: Contribution to journalJournal articlepeer-review

When training deep learning models for least-squares regression, we cannot expect that the training error residuals of the final model, selected after a fixed training time or based on performance on a hold-out data set, sum to zero. This can introduce a systematic error that accumulates if we are interested in the total aggregated performance over many data points (e.g., the sum of the residuals on previously unseen data). We suggest adjusting the bias of the machine learning model after training as a default post-processing step, which efficiently solves the problem. The severeness of the error accumulation and the effectiveness of the bias correction are demonstrated in exemplary experiments.

Original languageEnglish
JournalKI - Kunstliche Intelligenz
Number of pages8
Publication statusAccepted/In press - 2023

Bibliographical note

Funding Information:
The authors acknowledge support by the Villum Foundation through the project Deep Learning and Remote Sensing for Unlocking Global Ecosystem Resource Dynamics (DeReEco, project number 34306) and the Pioneer Centre for AI, DNRF grant number P1.

Publisher Copyright:
© 2023, The Author(s).

    Research areas

  • Bias correction, Deep learning, Regression

ID: 347311460