Abstract
Researchers in computational psycholinguistics frequently use linear models to study time series data generated by human subjects. However, time series may violate the assumptions of these models through temporal diffusion, where stimulus presentation has a lingering influence on the response as the rest of the experiment unfolds. This paper proposes a new statistical model that borrows from digital signal processing by recasting the predictors and response as convolutionally-related signals, using recent advances in machine learning to fit latent impulse response functions (IRFs) of arbitrary shape. A synthetic experiment shows successful recovery of true latent IRFs, and psycholinguistic experiments reveal plausible, replicable, and fine-grained estimates of latent temporal dynamics, with comparable or improved prediction quality to widely-used alternatives.- Anthology ID:
- D18-1288
- Volume:
- Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing
- Month:
- October-November
- Year:
- 2018
- Address:
- Brussels, Belgium
- Venue:
- EMNLP
- SIG:
- SIGDAT
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 2679–2689
- Language:
- URL:
- https://aclanthology.org/D18-1288
- DOI:
- 10.18653/v1/D18-1288
- Cite (ACL):
- Cory Shain and William Schuler. 2018. Deconvolutional Time Series Regression: A Technique for Modeling Temporally Diffuse Effects. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 2679–2689, Brussels, Belgium. Association for Computational Linguistics.
- Cite (Informal):
- Deconvolutional Time Series Regression: A Technique for Modeling Temporally Diffuse Effects (Shain & Schuler, EMNLP 2018)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/D18-1288.pdf
- Code
- coryshain/dtsr
- Data
- Natural Stories