Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sunsided/arima
Experiments on AR(p)/MA(q) processes in MATLAB.
https://github.com/sunsided/arima
arima arima-model artificial-intelligence matlab signal-processing statistics
Last synced: 10 days ago
JSON representation
Experiments on AR(p)/MA(q) processes in MATLAB.
- Host: GitHub
- URL: https://github.com/sunsided/arima
- Owner: sunsided
- Created: 2022-07-23T15:21:43.000Z (over 2 years ago)
- Default Branch: master
- Last Pushed: 2023-08-26T13:33:41.000Z (about 1 year ago)
- Last Synced: 2024-10-11T02:31:13.768Z (27 days ago)
- Topics: arima, arima-model, artificial-intelligence, matlab, signal-processing, statistics
- Language: MATLAB
- Homepage:
- Size: 231 KB
- Stars: 4
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# AR/MA model experiments
AR(p) and MA(q) model exploration in MATLAB using custom sample autocorrelation and sample partial autocorrelation implementations.
## AR(2) example
The following picture shows ACF and PACF plots for an AR(2) process with the parameters
```
y(0) = -42
y(t) = 0.7 y(t-1) + 0.2 y(t-2) + WN(my=0, sigma=1)
```
The ACF plot shows dampened sinusoidal behavior, indicating an AR(p) process, while the PACF shows no significant value after lag p=2, indicating AR(2).
## MA(1) example
The following picture shows ACF and PACF plots for an MA(1) process with the parameters
```
y(0) = epsilon(0) - 42
y(t) = 0.8 epsilon(t-1) - 42
```
The PACF plot shows dampened sinusoidal behavior, indicating an MA(q) process, while the ACF shows no significant value after lag p=1, indicating MA(1).
## 95% confidence intervals
The confidence intervals of the ACF and PACF plots are set to `± 1.96/√N`. Here `N` is the number of observations and `1.96` is the number of standard deviations 95% of the correlation values are expected lie within in under the assumption of the null hypothesis of no correlation.