In this paper, we study small deviation theorems of moving average for arbitrary integral random sequences. In suitable restrict conditions, we have lim infNh -1(N)∑N+h(N)i=NX iα(r(ω)), lim supNh -1(N)∑N+h(N)i=NX iβ(r(ω))
In this paper,we study small deviation theorems of moving average for arbitrary integral random sequences. In suitable restrict condition,we have\ lim infN h -1(N)∑N+h(N)i=NX i≥α(c), lim supN h -1(N)∑N+h(N)i=N X i≤β(c)
This paper presents a frequency domain sufficient condition of obtaining the stable moving average (MA) process by using the multidimensional Gevers Wouters algorithm, and gives an application to the construction of the autoregressive moving average (ARMA) innovation model, where a time domain sufficient condition to guarantee its moving average polynomial matrix to be stable is given.
The annual average 0 cm ground surface temperature of the stations along the Qinghai-Xizang railwaycorrelated well to each other,especially the correlationof 10-year running mean average sequences of them,which is as high as 0.92.According to the results,the 0 cm average ground surface temperature series along the Qinghai-Xizang railway from 1961 to 2003 are constructed.
Using summer temperature data in Xinjiang from 1959 to 2000 and the methods of running mean, maximum entropy spectrum, empirical orthogonal function and linear regression, the basic features of summer temperature change in the Tianshan mountainous region for the recent 42 years were analyzed and compared with those in the Northern Xinjiang and the Southern Xinjiang.
During 1990s, the yearly mean temperature of the warmest year (1998) or the temperature of the 5-year running mean almost reached to or was warmer than that in 1946 which was the warmest year and that of the 5-year running mean during the previous warm period.
Air-temperatures of representative ocean stations (Dalian, Longkou, Qianliyan and Lianyungang) along the coast of North China throughout 1976-1985 and the sea surface temperatures of Yellow Sea during 1976-1986 have been taken by ten days, and their running mean has been obtained for ten days, twenty days and sixty days respectively.
To reveal the spatial and temporal distribution variation characteristics of drought and flood of China,the drought or flood data of 100 stations in China from 1470 to 1996 were analyzed by using statistical analysis methods,such as overlapping average,EOF expansion,spectral analysis,and so on.
Two methods of signal smoothing were considered: square-wave (moving average) and triangular (weighted average) filters.
On Guaranteed Estimation of the Spectral Density of an Autoregression-Moving Average Process
An estimate for the spectral density of a stationary autoregression-moving average process with a given mean-square accuracy is proposed.
When the interval used for the moving average is increased to 8-12 years, a quasi-22-year rotational period appears.
We firstly propose a method to fit the fractional autoregressive integrated moving average (FARIMA) models to four actual wireless traffic traces and a traffic prediction based on the upper probability limit.