SigMA: Path Signatures and Multi-head Attention for Learning Parameters in fBm-driven SDEs

By Xianglin Wu, Chiheb Ben Hammouda, Cornelis W. Oosterlee

Published 2026-03-14

Everscope rating

1711.8

Relevance to quantitative trading

8 / 10

Implementation complexity

6 / 10

Reproducibility

5 / 5

About this paper

Methodology: SigMA (Signature Multi-head Attention). Problem types: Regression, Parameter Estimation, Time Series Analysis, Multi-task Learning (joint multi-parameter inference).

arXiv:2512.15088 · Code · Paper rankings

Open the interactive Everscope explorer for full analysis, charts, and paper battles.