Multi-Task Dynamical Systems
Not on the map. Why?
Abstract (arXiv)
Time series datasets are often composed of a variety of sequences from the same domain, but from different entities, such as individuals, products, or organizations. We are interested in how time series models can be specialized to individual sequences (capturing the specific characteristics) while still retaining statistical power by sharing commonalities across the sequences. This paper describes the multi-task dynamical system (MTDS); a general methodology for extending multi-task learning (MTL) to time series models. Our approach endows dynamical systems with a set of hierarchical latent variables which can modulate all model parameters. To our knowledge, this is a novel development of MTL, and applies to time series both with and without control inputs. We apply the MTDS to motion-capture data of people walking in various styles using a multi-task recurrent neural network (RNN), and to patient drug-response data using a multi-task pharmacodynamic model.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 31 of its 102 references that are in the map. First posted on arXiv 8 Oct 2022.
Builds on
31 of 102 references.
References · 31
- 2014Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translationcited by 4,016
- 2017DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networkscited by 1,004
- 2015Scheduled Sampling for Sequence Prediction with Recurrent Neural Networkscited by 779
- 2015Generating Sentences from a Continuous Spacecited by 607
- 2017MoCoGAN: Decomposing Motion and Content for Video Generationcited by 523
- 2015Importance Weighted Autoencoderscited by 454
- 2017On human motion prediction using recurrent neural networkscited by 336
- 2016Variational Lossy Autoencodercited by 175
- 2017Decomposing Motion and Content for Natural Video Sequence Predictioncited by 168
- 2016Deep Variational Bayes Filters: Unsupervised Learning of State Space Models from Raw Datacited by 154
- 2017Unsupervised Learning of Disentangled Representations from Videocited by 138
- 2018Learning to Decompose and Disentangle Representations for Video Predictioncited by 120
19 more
- 2018Disentangled Sequential Autoencodercited by 117
- 2018QuaterNet: A Quaternion-based Recurrent Model for Human Motioncited by 108
- 2017Variational Sequential Monte Carlocited by 93
- 2017Unsupervised Learning of Disentangled and Interpretable Representations from Sequential Datacited by 92
- 2017Filtering Variational Objectivescited by 91
- 2017Recurrent Environment Simulatorscited by 68
- 2017Learning Human Motion Models for Long-term Predictionscited by 67
- 2017Auto-Encoding Sequential Monte Carlocited by 64
- 2014Variational Recurrent Auto-Encoderscited by 53
- 2017Z-Forcing: Training Stochastic Recurrent Networkscited by 51
- 2016Personalized Risk Scoring for Critical Care Prognosis using Mixtures of Gaussian Processescited by 19
- 2019Disentangled State Space Representationscited by 14
- 2016A Framework for Individualizing Predictions of Disease Trajectories by Exploiting Multi-Resolution Structurecited by 9
- 2019Patient-Specific Effects of Medication Using Latent Force Models with Gaussian Processescited by 6
- 2017Treatment-Response Models for Counterfactual Reasoning with Continuous-time, Continuous-valued Interventionscited by 4
- 2018Clustering Time Series with Nonlinear Dynamics: A Bayesian Non-Parametric and Particle-Based Approachcited by 4
- 2019Multi-Task Time Series Analysis applied to Drug Response Modellingcited by 2
- 2016Input-Output Non-Linear Dynamical Systems applied to Physiological Condition Monitoringcited by 1
- 2019Updating Variational Bayes: Fast sequential posterior inferencecited by 1
Built on by
Citing papers per year: 2023: 1.
1 paper.