Work

The credibility transformer

Year: 2025

Type: article

Abstract: Inspired by the large success of Transformers in Large Language Models, these architectures are increasingly applied to tabular data. This is achieved by embedding tabular data into low-dimensional Euclidean spaces resulting in similar structures as time-series data. We introduce a novel credibility mechanism to this Transformer architecture. This credibility mechanism is based on a special token that should be seen as an encoder that consists of a credibility weighted average of prior information and observation based information. We demonstrate that this novel credibility mechanism is very beneficial to stabilize training, and our Credibility Transformer leads to predictive models that are superior to state-of-the-art deep learning models.

Source: European Actuarial Journal

DOI: 10.1007/s13385-025-00413-y

Authors: Ronald Richman, Salvatore Scognamiglio, Mario V. Wüthrich

Institutions: University of the Witwatersrand, Parthenope University of Naples, ETH Zurich

Language: English

Topic: Time Series Analysis and Forecasting

Subfield: Signal Processing

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Estimated APC paid: $2,290

Funders: University of New South Wales, Ewha Womans University, Stockholms Universitet