Emotion Forecasting: A Transformer-Based Approach
The stability of multivariate time-series forecasting improved when HMM preprocessing was combined with a transformer model, as opposed to other time-series forecasting methods (eg, recurrent neural network and long short-term memory), leveraging the attention mechanisms to capture longer time dependencies and gain interpretability. We showed the potential to assess the emotional state of a patient and the scores of psychiatric questionnaires from passive variables in advance. This allows real-time monitoring of patients and hence better risk detection and treatment adjustment.
Authors
- David Ramírez (ORCID: https://orcid.org/0000-0002-6883-7742)
- Pablo M. Olmos (ORCID: https://orcid.org/0000-0003-4276-5178)
- Antonio Artés-Rodrı́guez (ORCID: https://orcid.org/0000-0001-6540-7109)
- Jorge López‐Castromán (ORCID: https://orcid.org/0000-0003-0618-0690)
- Leire Paz-Arbaizar (ORCID: https://orcid.org/0009-0002-9741-2652)
Institutions
- Universidade de Santiago de Compostela (ES)
- Hospital General Universitario Gregorio Marañón (ES)
- Centro de Investigación Biomédica en Red de Salud Mental (ES)
- Evidence Based Research (United States) (US)
- Universidad Carlos III de Madrid (ES)
- Université de Nîmes (FR)
Publication Details
- Journal
- Journal of Medical Internet Research
- Published
- 2025-01-16
- DOI
- https://doi.org/10.2196/63962
- Citations
- 7
- Primary Topic
- Emotion and Mood Recognition
- Type
- article
- Field-Weighted Citation Impact
- 14.83