State Space Model
Definition
A state space model explains temporal observations through a hidden, changing system state.
A state space model describes a dynamic system via a latent state that evolves over time and generates observations.
Simply explained
You see measurements, but not everything behind them. The model estimates a hidden state from these and tracks its development.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A forecasting system models hidden demand states from a sequence of noisy measurements.
Limits and pitfalls
- Non-identifiable states, incorrect dynamic assumptions, and drift can make forecasts unreliable.
Origin & history
The modern state space representation originated in 20th-century control and estimation theory; newer sequential networks use structured SSMs.
Use in mid-sized companies
- Possible application: A forecasting system models hidden demand states from a sequence of noisy measurements.
Frequently asked questions
What does state space model mean?
A state space model describes a dynamic system via a latent state that evolves over time and generates observations. What should you watch out for? Non-identifiable states, incorrect dynamic assumptions, and drift can make forecasts unreliable.
Sources
- A New Approach to Linear Filtering and Prediction Problems · Kalman, A New Approach to Linear Filtering and Prediction Problems
Source checked on 10.10.2026. Editorial application examples are hypothetical; no evidence of real customer results.
Editorially reviewed on: