stationary stochastic process
Học thuậtThân thiện
Definition
- Noun:
- A stationary stochastic process is a specific type of stochastic (random) process. Its defining property is that its statistical characteristics do not change over time or across the index parameter. More precisely, the joint probability distribution of any collection of random variables from the process remains invariant when shifted in time or index.
Usage
- The term is used in technical fields like statistics, econometrics, signal processing, and time series analysis to describe systems or data whose underlying randomness has stable, time-invariant properties.
- It describes a theoretical model or an observed sequence where the mean, variance, and autocorrelation structure are constant over time.
Examples
- Noun:
- White noise is a fundamental example of a stationary stochastic process.
- For reliable forecasting, many time series models assume the data is generated by a stationary stochastic process.
- The researcher tested whether the economic data could be modeled as a stationary stochastic process.
Advanced Usage
- "Weakly/wide-sense stationary process": A less restrictive condition requiring only that the mean is constant and the autocovariance depends only on the time lag, not on the absolute time.
- Many practical applications rely on the assumption of a weakly stationary stochastic process.
- "Strictly stationary process": The stronger condition defined above, where the entire distribution is invariant to time shifts.
- Proving a process is a strictly stationary stochastic process is often more difficult than proving weak stationarity.
Variants and Related Words
- Stationarity (n): The property of being stationary.
- The stationarity of the process is a key assumption.
- Non-stationary process (n): A stochastic process whose statistical properties change over time.
- Many real-world financial time series are non-stationary processes.
Synonyms
- Stationary process (The term "stochastic" is often implied in technical contexts).
- Time-invariant random process.
Related Concepts
- Ergodic process: A stationary stochastic process where time averages equal ensemble averages, allowing properties to be inferred from a single sample path.
- Autocorrelation function: A key tool for analyzing the properties of a stationary stochastic process.
Noun
- a stochastic process in which the distribution of the random variables is the same for any value of the variable parameter