Supervision : Ludovic DENOYER

Co-supervision : LAMPRIER Sylvain

Inferring and Predicting Dynamic Representations for Structured Temporal Data

Temporal data constitute a large part of data collected digitally. Predicting their next values is an important and challenging task in domains such as climatology, optimal control, or natural language processing. Standard statistical methods are based on linear models and are often limited to low-dimensional data. We instead use deep learning methods capable of handling high-dimensional structured data and leverage large quantities of examples.
In this thesis, we are interested in latent variable models. Contrary to autoregressive models that directly use past data to perform prediction, latent models infer low-dimensional vectorial representations of data on which prediction is performed.
First, we propose a structured latent model for spatiotemporal data forecasting. Next, we focus on predicting data distributions, rather than point estimates, through a diachronic language modeling task. Finally, we propose a stochastic prediction model applied to video prediction.

2017-2021 Publications

• 2021
  • E. Delasalles, S. Lamprier, L. Denoyer : “Deep dynamic neural networks for temporal language modeling in author communities”, Knowledge and Information Systems (KAIS), vol. 63 (3), pp. 733-757, (Springer) (2021)
• 2020
  • E. Delasalles : “Déduire et prédire des représentations dynamiques pour des données temporelles structurées”, thesis, phd defence 06/30/2020, supervision Denoyer, Ludovic, co-supervision : Lamprier, Sylvain (2020)
  • J.‑Y. Franceschi, E. Delasalles, M. Chen, S. Lamprier, P. Gallinari : “Stochastic Latent Residual Video Prediction”, Proceedings of the 37^th International Conference on Machine Learning, vol. 119, Proceedings of Machine Learning Research, Vienne, Austria, pp. 3233-3246, (PMLR) (2020)
• 2019
  • E. Delasalles, S. Lamprier, L. Denoyer : “Dynamic Neural Language Models”, ICONIP 2019 - 26^th International Conference on Neural Information Processing, vol. 11955, Lecture Notes in Computer Science, Sydney, Australia, pp. 282-294 (2019)
  • E. Delasalles, A. Ziat, L. Denoyer, P. Gallinari : “Spatio-temporal neural networks for space-time data modeling and relation discovery”, Knowledge and Information Systems (KAIS), vol. 61 (3), pp. 1241-1267, (Springer) (2019)
  • E. Delasalles, S. Lamprier, L. Denoyer : “Learning Dynamic Author Representations with Temporal Language Models”, 2019 IEEE International Conference on Data Mining (ICDM), Beijing, China, pp. 120-129, (IEEE) (2019)
• 2018
  • E. Delasalles, S. Lamprier, L. Denoyer : “Apprentissage de l’évolution langagière dans des communautés d’auteurs”, 15^e COnférence en Recherche d'Informations et Applications - CORIA 2018, Rennes, France (2018)
• 2017
  • A. Ziat, E. Delasalles, L. Denoyer, P. Gallinari : “Spatio-Temporal Neural Networks for Space-Time Series Forecasting and Relations Discovery”, 2017 IEEE International Conference on Data Mining (ICDM), 2017 IEEE International Conference on Data Mining (ICDM), La Nouvelle Orléans, LA, United States, pp. 705-714, (IEEE) (2017)