基於時間序列模型的預測
阿新 • • 發佈：2019-01-07
Statespace Model Results
==========================================================================================
Dep. Variable: riders No. Observations: 114
Model: SARIMAX(0, 1, 0)x(0, 1, 1, 12) Log Likelihood -504.683
Date: Tue, 20 Jun 2017 AIC 1013.365
Time: 05:37:32 BIC 1018.838
Sample: 01-31-1960 HQIC 1015.586
- 06-30-1969
Covariance Type: opg
==============================================================================
coef std err z P>|z| [0.025 0.975]
------------------------------------------------------------------------------
ma.S.L12 -0.6937 0.118 -5.867 0.000 -0.925 -0.462
sigma2 1185.5644 183.539 6.459 0.000 825.834 1545.295
===================================================================================
Ljung-Box (Q): 43.16 Jarque-Bera (JB): 2.01
Prob(Q): 0.34 Prob(JB): 0.37
Heteroskedasticity (H): 1.49 Skew: 0.28
Prob(H) (two-sided): 0.25 Kurtosis: 3.39
===================================================================================
Warnings:
[1] Covariance matrix calculated using the outer product of gradients (complex-step).
Residuals Summary:
0
count 114.000000
mean 0.920410
std 72.659634
min -214.018001
25% -22.898384
50% -5.576031
75% 18.514757
max 648.000000