We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Methodology

Title: CovNet: Covariance Networks for Functional Data on Multidimensional Domains

Abstract: Covariance estimation is ubiquitous in functional data analysis. Yet, the case of functional observations over multidimensional domains introduces computational and statistical challenges, rendering the standard methods effectively inapplicable. To address this problem, we introduce Covariance Networks (CovNet) as a modeling and estimation tool. The CovNet model is universal -- it can be used to approximate any covariance up to desired precision. Moreover, the model can be fitted efficiently to the data and its neural network architecture allows us to employ modern computational tools in the implementation. The CovNet model also admits a closed-form eigen-decomposition, which can be computed efficiently, without constructing the covariance itself. This facilitates easy storage and subsequent manipulation in the context of the CovNet. Moreover, we establish consistency of the proposed estimator and derive its rate of convergence. The usefulness of the proposed method is demonstrated by means of an extensive simulation study.
Subjects: Methodology (stat.ME); Statistics Theory (math.ST); Machine Learning (stat.ML)
Cite as: arXiv:2104.05021 [stat.ME]
  (or arXiv:2104.05021v1 [stat.ME] for this version)

Submission history

From: Soham Sarkar [view email]
[v1] Sun, 11 Apr 2021 14:40:41 GMT (1603kb,D)

Link back to: arXiv, form interface, contact.