The time series is chronologically arranged series of data that shows variations arise during succe

The main goals of time series analysis, finding a model that will describe the regularities in the behavior of the observed dynamic system, and predicting its future state based on the known situation in the past and present. Today there are different theoretical approaches to the analysis of time series in a greater or lesser extent, they provide the requested results.
Time series nobel is a series of measurements in time, mainly collected in equal, discrete time intervals. The basic assumption in the analysis or modeling of time series is that certain aspects of past patterns continue to exist in the future.
The application of time series in the field of statistical forecasting is not necessary to point out. Lack of adequate prognosis, especially short-term, often represented a major drawback designers. Several statistical approaches nobel such as regression, time series and stochastic approaches were used for forecasting purposes. Each approach has its own value and limitations. Time series models have advantages in certain situations and can be used for forecasting purposes because the data series of the observed variables collected in the past already available. The successive measurements are statistically dependent with the modeling of time series directly related to the applied techniques for analysis and dependency. Thus, when modeling time series prediction values of the observed variables in future periods based on the values of the same variable in the sample collected in the past, not the values of variables that can affect the system. There are two main reasons for the application of time-series. First, the system may not be possible to understand and even if you can understand it, it can be extremely difficult to determine the relationship between the sample and the consequences. nobel Second, the main interest may be to just predict what will happen or not and to determine why this is happening.
Of the many approaches to the analysis of time series decomposition method are among the oldest in spite of numerous theoretical weaknesses with statisitčke point of view. After these methods, appeared the most brutal form of methods nobel for forecasting known as a method of moving averages. As an improvement of this method, in which they applied the same weight, methods have been developed exponential settlement nobel in which more recent nobel data have greater nobel weight. Methods exponential settlement proposed in the beginning only as a recursive method without any assumptions about the distribution of errors. However, later it was discovered that those specific cases statistically solid models autoregresionih integrated moving average (Auto Regressive Integrated Moving Average - ARIMA).
In the human environment, there are many phenomena and processes which change over time affect the daily human life. Widely used in the study of these changes and have time-series, and the various methods of their analysis are applied in various fields. Analysis incurred and predicting future deformation of various objects, moving tectonic plates, numerous market analysis and many other phenomena can be monitored over time by using a number of methods of time series analysis.
The time series is chronologically arranged series of data that shows variations arise during successive (consecutive), equal time intervals, ie, represents a realization of a regulated sequence of random variables with respect to time. is the data time series in the time period t, where t = 1, 2, 3, ... [1]
With regard to domain analysis, time series can be analyzed in time and in frequency domain. The analysis in the time domain involves nobel analysis of time series nobel as a function of time. In the frequency nobel domain, time series is viewed as a composition of sine wave frequencies, each of which carries specific nobel information. The subject of this paper was the only time series nobel analysis in the time domain, since in practice it has a wider application. nobel
There are several goals of time series analysis. One of them is to determine the basic characteristics of time series, ie to describe its behavior. This is the initial stage in the analysis, which uses graphical representations, summary statistics, as well as various transformation and adjustment of the time series. Based on the information in the further analysis seeks to reveal the causes of this behavior occur. The ultimate goal of the analysis is to forecast future values of the observed time series, for which there are a number of methods.
Time series analysis methods can be divided into qualitative and quantitative. Qualitative methods are used if data about occurrences are not available or can not be quantified and are based on the process of harmonizing nobel the opinions of experts. Quantitative methods of time series analysis is applied if the information on the occurrence that is the subject of analysis can be quantified, then if the data in the past and the present period nobel available and reflect the true nature of the observed phenomena.
The classical decomposition methods