Physics > Data Analysis, Statistics and Probability
[Submitted on 11 Dec 2014 (this version), latest version 7 Mar 2016 (v3)]
Title:Analog Forecasting with Dynamics-Adapted Kernels
View PDFAbstract:Analog forecasting is a non-parametric technique introduced by Lorenz in 1969 which predicts the evolution of states of a dynamical system (or observables defined on the states) by following the evolution of the sample in a historical record of observations which most closely resembles the current initial data. Here, we introduce a suite of forecasting methods which improve traditional analog forecasting by combining ideas from state-space reconstruction for dynamical systems and kernel methods developed in harmonic analysis and machine learning. The first improvement is to augment the dimension of the initial data using Takens' delay-coordinate maps to recover information in the initial data lost through partial observations. Then, instead of using Euclidean distances between the states, weighted ensembles of analogs are constructed according to similarity kernels in delay-coordinate space, featuring an explicit dependence on the dynamical vector field generating the data. The eigenvalues and eigenfunctions of associated kernel operators define diffusion coordinates and diffusion distances on the data manifold, improving the identification of skillful analogs compared to the Euclidean distance. Moreover, forecasts based on kernel-weighted ensembles have significantly higher skill than the conventional approach following a single analog. We discuss alternative approaches for kernel-weighted ensemble forecasting based on the Nyström method for out-of-sample extension of functions, as well as multiscale methods based on Laplacian pyramids. We illustrate these techniques in applications to forecasting in a low-order deterministic model for atmospheric dynamics with chaotic metastability, and interannual-scale forecasting in the North Pacific sector of a comprehensive climate model.
Submission history
From: Zhizhen Zhao [view email][v1] Thu, 11 Dec 2014 21:34:40 UTC (138 KB)
[v2] Sat, 30 Jan 2016 23:36:31 UTC (779 KB)
[v3] Mon, 7 Mar 2016 23:21:25 UTC (779 KB)
Current browse context:
physics.data-an
Change to browse by:
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.