Gaussian processes for time-series modelling

By Stephen J. Roberts, Michael A. Osborne, Mark Ebden, Steven Reece, Neale P. Gibson, Suzanne Aigrain. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2013

623 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Signal Processing

View Paper: https://www.semanticscholar.org/paper/64e615720384525dfa317605e154ee259ba1013d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Signal Processing

Citations in the corpus

Modeling Long- and Short-Term Temporal Patterns with Deep Neural Networks (2017, methodology): “[11, 29] • VAR-MLP is the model proposed in [36] that combines Multilayer Perception (MLP) and autoregressive model.”
Temporal pattern attention for multivariate time series forecasting (2018, methodology): “– GP: Gaussian process model [Frigola-Alcade(2015), S. Roberts and Aigrain(2011)].”
Identifying Platform Effects in Social Media Data (2016, methodology): “We also tried Gaussian Process (GP) regression (Rasmussen and Williams 2005; MacDonald, Ranjan, and Chipman 2015), as it is able to capture temporal dependencies (Roberts et al. 2012).”
Sparse Multi-Output Gaussian Processes for Medical Time Series Prediction (2017, methodology): “rocesses (GPs) are useful approaches for time series analysis because they can naturally capture irregular time series observations and estimate prediction uncertainties in a probabilistic framework (Roberts et al., 2012). For these reasons, GPs have been applied to the analysis of medical time seri”
A Gaussian process framework for modelling stellar activity signals in radial velocity data (2015, methodology): “This process is known as GP regression, and is described in more detail in textbooks such as Bishop (2006) and Rasmussen & Williams (2006); alternatively, see Roberts et al. (2013) for an accessible, tutorial-style introduction to GP regression.”
Graph Neural Networks for Intelligent Transportation Systems: A Survey (2023, methodology): “They compared their model with baselines for both single-step (autoregressive, VAR-MLP [38], Gaussian Process [39], RNN-GRU, LSTNet [40], TPA-LSTM [41]) and multi-step (DCRNN, STGCN, Graph WaveNet, ST-MetaNet, GMAN, and MRA-BGCN [42]) forecasting, and reported the overall competence of their…”
Time Series Data Augmentation for Deep Learning: A Survey (2020, methodology): “Gaussian Processes (GPs) [Rasmussen and Williams, 2005] are well-known Bayesian non-parametric models suitable for time series analysis [Roberts et al. , 2013].”
MOrdReD: Memory-based Ordinal Regression Deep Neural Networks for Time Series Forecasting (2018, methodology): “Consequently, the time series literature has increasingly been dominated by models such as the Gaussian Process [5, 6, 7, 8, 9] which enable the inference of full predictive distributions in a principled manner, even in the presence of little data.”
Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks (2020, neutral): “• GP: A Gaussian Process time series model [6, 16].”
GP-VAE: Deep Probabilistic Time Series Imputation (2019, background): “Gaussian processes are well suited for time series modeling [41] and oﬀer many advantages, such as data-eﬃciency and calibrated posterior probabilities.”
Remarks on multi-output Gaussian process regression (2018, background): “, small- or largescale regression [4, 5, 6], dimensionality reduction [7], Bayesian optimization [8], uncertainty quantification [9] and time-series analysis [10].”
Text embedding models can be great data engineers (2025, unknown): “This process involves a series of operations, such as: Here, f imp denotes missing-value interpolation (e.g. cubic-spline [28] or Gaussian-process [43]), f norm represents normalization (min–max or z -score), and ϕ extracts M handcrafted features (often M ≫ TFK ; e.g. TSFEL yields M > 9000 [5]).”