Skip to contentScholarGate
LibraryBookshelfDeskReview StudioAssistant
Sign in
On this page
IntuitionHow it worksWhen to use itStrengths & limitationsCommon pitfallsApplicationsFrequently asked🔒 Read the full methodSourcesRelated methods
Cite this pageSpotted an issue on this page? Report or suggest a fix →
Home›Bayesian›Time Series Kalman Filter
Bayesian methodsBayesian / computational

Time Series Kalman Filter

Kalman Filter for Time Series State-Space Models · Also known as: state-space Kalman filter, structural time series filter, Kalman smoother for time series, time series state-space filter

The time series Kalman filter applies the Kalman filtering and smoothing algorithm within a state-space representation of time series models. It recursively extracts unobserved components — trend, seasonality, cycles, and irregular noise — from observed data, providing optimal filtered and smoothed state estimates together with their uncertainty, and enabling exact likelihood evaluation for parameter estimation.

ScholarGate
  1. Bayesian methods
  2. v1
  3. 2 Sources
  4. PUBLISHED
Cite this page →
Tools & resources
Download slides
Learn & explore

Read the full method

Members only

Sign in with a free account to read this section.

Sign in

Method map

The neighbourhood of related methods — select a node to explore.

Time Series Kalman Filter
Bayesian RegressionDynamic Bayesian NetworkKalman FilterParticle FilterSequential Monte CarloTime series Bayesian inf…Time series particle fil…

When to use it

Use the time series Kalman filter when you want to decompose a time series into interpretable unobserved components (trend, seasonal, cycle, irregular), when observations are missing or irregularly spaced, when you need an exact likelihood for maximum likelihood or Bayesian parameter estimation, or when the data-generating process can be plausibly cast in linear-Gaussian state-space form. It is particularly powerful for structural time series models and state-space econometrics. Do not use it when the system dynamics are strongly nonlinear or the error distributions are non-Gaussian without modification — in those cases use an Extended Kalman Filter, Unscented Kalman Filter, or particle filter. Also avoid it when the series is short relative to the number of unknown parameters, as the likelihood surface can be flat.

Strengths & limitations

Strengths
  • Handles missing values and irregular spacing with no modification — just skip the update step when an observation is absent.
  • Provides exact Gaussian log-likelihood evaluation, enabling straightforward maximum likelihood or Bayesian estimation of hyperparameters.
  • Decomposes series into interpretable unobserved components (trend, seasonal, irregular) with associated uncertainty bands.
  • Computationally efficient and recursive — O(n) in the number of time points for fixed model order.
  • Unified framework that nests ARIMA, exponential smoothing, and structural time series models as special cases.
Limitations
  • Assumes linear state equations and Gaussian disturbances — violated by count data, heavy-tailed noise, or nonlinear dynamics without extension.
  • Requires specification of system matrices (T, Z, Q, H) whose unknown parameters must be estimated, which can be difficult in high-dimensional state spaces.
  • Likelihood surface for hyperparameter estimation can be multimodal or flat, causing optimisation to find poor local solutions.
  • Diffuse initialisation for nonstationary components requires careful handling, especially for short series.

Frequently asked

What is the difference between filtering and smoothing?

Filtering produces the state estimate a_t using only observations up to time t, making it suitable for real-time applications. Smoothing (e.g., the Kalman smoother) re-runs a backward pass after the full forward filter and uses all observations to refine each state estimate, reducing uncertainty. Use smoothed estimates for historical component extraction; use filtered estimates for online forecasting.

How are hyperparameters (Q, H) estimated?

Unknown variance parameters are typically estimated by maximising the Gaussian log-likelihood formed from the innovations and innovation variances produced by the Kalman filter. This can be done with numerical optimisation (e.g., L-BFGS-B) or, in a fully Bayesian framework, by placing priors on the variances and sampling with MCMC.

Can the Kalman filter handle non-Gaussian or count data?

Not directly. For non-Gaussian observation distributions (Poisson counts, binary data), the exact Kalman filter no longer applies. Extensions include the Extended Kalman Filter (linearisation), the Unscented Kalman Filter (sigma-point approximation), and particle filters (sequential Monte Carlo), all of which preserve the state-space structure while relaxing the Gaussian assumption.

How does the state-space approach relate to ARIMA models?

Every ARIMA(p,d,q) model can be written in state-space form and estimated via the Kalman filter. The state-space representation is more general, however: it accommodates structural components, multiple observation series, time-varying coefficients, and missing data within the same recursive algorithm.

What is diffuse initialisation and why does it matter?

When the initial state is nonstationary (e.g., a random-walk trend with unknown starting level), the prior variance is effectively infinite. Diffuse initialisation handles this rigorously by using an exact diffuse prior, so that the first few innovations are computed correctly without arbitrarily large or hard-coded initial covariance values. Using a naive large P_0 can distort the likelihood and parameter estimates for short series.

Sources

  1. Durbin, J. & Koopman, S. J. (2012). Time Series Analysis by State Space Methods (2nd ed.). Oxford University Press. ISBN: 978-0199641178
  2. Harvey, A. C. (1989). Forecasting, Structural Time Series Models and the Kalman Filter. Cambridge University Press. ISBN: 978-0521321969

How to cite this page

ScholarGate. (2026, June 3). Kalman Filter for Time Series State-Space Models. ScholarGate. https://scholargate.app/en/bayesian/time-series-kalman-filter

Related methods

Bayesian RegressionDynamic Bayesian NetworkKalman FilterParticle FilterSequential Monte CarloTime series Bayesian inference

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

  • Bayesian RegressionBayesian↔ compare
  • Dynamic Bayesian NetworkBayesian↔ compare
  • Kalman FilterBayesian↔ compare
  • Particle FilterBayesian↔ compare
  • Sequential Monte CarloBayesian↔ compare
  • Time series Bayesian inferenceBayesian↔ compare
Compare side by side →

Referenced by

Time series particle filter

Similar methods

State Space ModelKalman Filter with Missing DataKalman Filter with Measurement ErrorKalman Filter (Finance)Kalman FilterSpatial Kalman FilterTime series Bayesian inferenceStructural Time Series Model

Related reference concepts

Hidden Markov ModelsData AssimilationEM AlgorithmGaussian Process ModelsTime-Series Models • Dynamic Quantile Regressions • Dynamic Treatment Effect Models • Diffusion Processes • State Space ModelsLatent Variable and Mixture Models

Spotted an issue on this page? Report or suggest a fix →

ScholarGate — Time Series Kalman Filter (Kalman Filter for Time Series State-Space Models). Retrieved 2026-07-21 from https://scholargate.app/en/bayesian/time-series-kalman-filter · Dataset: https://doi.org/10.5281/zenodo.20539026
Quick facts
Originator
Rudolf E. Kalman; systematised for time series by Andrew Harvey and Søren Johansen
Year
1960; time series state-space framework formalised 1980s–1990s
Type
recursive Bayesian filter / state-space smoother
DataType
univariate or multivariate time series, equally or irregularly spaced
Subfamily
Bayesian / computational
Related methods
Bayesian RegressionDynamic Bayesian NetworkKalman FilterParticle FilterSequential Monte CarloTime series Bayesian inference
ScholarGate

A content-first reference library for research methods — what each one is, how it works, and where it comes from.

Open data (CC-BY)

Explore

  • Library
  • Search the library…
  • Browse by field
  • Fields
  • Journey
  • Compare
  • Which method?

Reference

  • Subjects
  • Atlas
  • Glossary
  • Methodology
  • Philosophy

Your tools

  • Bookshelf
  • Desk
  • Chat

Company

  • About
  • Pricing
  • Contact
  • Suggest a method

Entries are compiled from published sources for reference. Verifying the accuracy and suitability of any information for your own use remains your responsibility.

© 2026 ScholarGate · A research-method reference library
  • Privacy
  • Cookies
  • Terms
  • Delete account