Time-series forecasting through recurrent topology
Abstract Time-series forecasting is a practical goal in many areas of science and engineering. Common approaches for forecasting future events often rely on highly parameterized or black-box models. However, these are associated with a variety of drawbacks including critical model assumptions, uncer...
Main Authors: | , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2024-01-01
|
Series: | Communications Engineering |
Online Access: | https://doi.org/10.1038/s44172-023-00142-8 |
_version_ | 1797355708459515904 |
---|---|
author | Taylor Chomiak Bin Hu |
author_facet | Taylor Chomiak Bin Hu |
author_sort | Taylor Chomiak |
collection | DOAJ |
description | Abstract Time-series forecasting is a practical goal in many areas of science and engineering. Common approaches for forecasting future events often rely on highly parameterized or black-box models. However, these are associated with a variety of drawbacks including critical model assumptions, uncertainties in their estimated input hyperparameters, and computational cost. All of these can limit model selection and performance. Here, we introduce a learning algorithm that avoids these drawbacks. A variety of data types including chaotic systems, macroeconomic data, wearable sensor recordings, and population dynamics are used to show that Forecasting through Recurrent Topology (FReT) can generate multi-step-ahead forecasts of unseen data. With no free parameters or even a need for computationally costly hyperparameter optimization procedures in high-dimensional parameter space, the simplicity of FReT offers an attractive alternative to complex models where increased model complexity may limit interpretability/explainability and impose unnecessary system-level computational load and power consumption constraints. |
first_indexed | 2024-03-08T14:15:01Z |
format | Article |
id | doaj.art-4411e458ff4b4cf2843e6b1bde8908b1 |
institution | Directory Open Access Journal |
issn | 2731-3395 |
language | English |
last_indexed | 2024-03-08T14:15:01Z |
publishDate | 2024-01-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Communications Engineering |
spelling | doaj.art-4411e458ff4b4cf2843e6b1bde8908b12024-01-14T12:25:03ZengNature PortfolioCommunications Engineering2731-33952024-01-013111010.1038/s44172-023-00142-8Time-series forecasting through recurrent topologyTaylor Chomiak0Bin Hu1Division of Translational Neuroscience, Department of Clinical Neurosciences, Hotchkiss Brain Institute, Alberta Children’s Hospital Research Institute, Cumming School of Medicine, University of CalgaryDivision of Translational Neuroscience, Department of Clinical Neurosciences, Hotchkiss Brain Institute, Alberta Children’s Hospital Research Institute, Cumming School of Medicine, University of CalgaryAbstract Time-series forecasting is a practical goal in many areas of science and engineering. Common approaches for forecasting future events often rely on highly parameterized or black-box models. However, these are associated with a variety of drawbacks including critical model assumptions, uncertainties in their estimated input hyperparameters, and computational cost. All of these can limit model selection and performance. Here, we introduce a learning algorithm that avoids these drawbacks. A variety of data types including chaotic systems, macroeconomic data, wearable sensor recordings, and population dynamics are used to show that Forecasting through Recurrent Topology (FReT) can generate multi-step-ahead forecasts of unseen data. With no free parameters or even a need for computationally costly hyperparameter optimization procedures in high-dimensional parameter space, the simplicity of FReT offers an attractive alternative to complex models where increased model complexity may limit interpretability/explainability and impose unnecessary system-level computational load and power consumption constraints.https://doi.org/10.1038/s44172-023-00142-8 |
spellingShingle | Taylor Chomiak Bin Hu Time-series forecasting through recurrent topology Communications Engineering |
title | Time-series forecasting through recurrent topology |
title_full | Time-series forecasting through recurrent topology |
title_fullStr | Time-series forecasting through recurrent topology |
title_full_unstemmed | Time-series forecasting through recurrent topology |
title_short | Time-series forecasting through recurrent topology |
title_sort | time series forecasting through recurrent topology |
url | https://doi.org/10.1038/s44172-023-00142-8 |
work_keys_str_mv | AT taylorchomiak timeseriesforecastingthroughrecurrenttopology AT binhu timeseriesforecastingthroughrecurrenttopology |