Abstract This paper analyzes China's crude steel output in the time series (January 2000 to 2012), to deal with the volatility of the trend of the time series in the order of the natural logarithm of differential and order season differential based on seasonal seasonal autoregressive moving average model (SARMA model). The results showed that the model fit better predicted data and actual data error is small http://www.cheapjordanbred11.com/, so that the model has a certain reference value.
-3871240.htm
Key words】 crude steel production; SARMA model; forecast
analysis method - SARMA model introduced the
SARMA model class time series model, the basic idea is: based on the variation of the variable itself, extrapolation mechanism to describe the changes in the time series. The premise is the time sequence is zero-mean stationary random process. For non-stationary series containing trend and seasonal jordan 11 2013, subject to appropriate phase-wise differential and seasonal differential eliminate the impact and then sequence analysis, the establishment of SARMA (p, d, q) (P, D, Q) s model. The general manifestation SARMA Model: φp (B) φp (Bs) (1-B) d (1-Bs) Dyt = θq (B) θ (Bs) μt. Wherein: yt for time series, μt random item, φp (Bs) and a non-season portion AR (p) portion, øp (BS) for the season ar (P) portion bred 11s online, (1-B) d is d order by period differential, (1-BS) D for D order-by-phase differential, θq (B) non-seasonal MA (q) part, Q (Bs) for seasonal MA (Q) part.
predictive model
1. data processing and analysis. The selected crude steel production data from January 2000 to December 2012
CEInet statistical database as objects, data from the statistical software eviews7.2. Figure 1 China's crude steel production line chart (denoted as q), we can see that China's crude steel production in 2008 showing a stable growth trend and seasonal fluctuations in recent years by the impact of the financial crisis, lower raw material costs The impact of domestic demand, lack of external demand, slow growth in crude steel production and frequent fluctuations.
log processing in order to reduce the sequence of changes in trends and heteroscedasticity, the original sequence, denoted by lq non-stationary sequence lq, the sequence dlq differential after smooth unit root test, the unit root test The results are shown in Table 1.
identification and establishment of the model. The sequence dlq in the lag period of 12, 24, 36, etc. autocorrelation coefficient significantly different from zero. The sequence to cycle 12 seasonal bred 11s price, first-order differential form a new sequence sdlq In order to consider this seasonal autocorrelation analysis is shown in Figure 2, sdlq lag a greater difference to 0, the lag period 24 lag season ACF and PACF has decayed to 0, so SARMA (p, d, q) (P, D, Q) 12 D = D = 1, p and q does not exceed 1, P and Q EXCEED 2. The
mainly by AIC and SC criteria to determine the most accurate order, taking into account the significant test parameters, residual significant test results. The
eviews7.2 for p = 0 michael kors watches outlet,1 q = 0,1, P = 1,2, and Q = 1,2 SARMA (P 1, Q) (P 1, Q) 12 model one by one projection model SARMA (0,1,0) (2,1,2) 12 most suitable. The model parameter estimation results are shown in Table 2.

reproduced please specify the source. Original Address: Related articles:

• http://www.sufc.co.za/forum/forum_posts.asp?TID=680684&PID=779294#779294
• http://www.bjopel.com/bbs/forum.php?mod=viewthread&tid=99668