### The advantages and disadvantages of time series forecasting method are urgent, online etc.

Time series analysis forecasting method has two characteristics:

①Time series analysis forecasting method is to predict the future development based on the past trend of the market, its premise is to assume that the past of the thing will be the same continuation into the future. The reality of things is the result of historical development, and the future of things is an extension of reality, the past and future of things are linked. The time series analysis of market forecasting is based on this continuous regularity of the development of objective things, the use of past historical data, through statistical analysis, and further speculation on the future development trend of the market. In market forecasting, the past of things will be the same continuation into the future, which means that the market future will not be a sudden jump in change, but gradual change.

The philosophical basis of the time series analysis and forecasting method is the basic view of material dialectics, that is, all things are development and change, the development and change of things in time has a continuity, the market phenomenon is also so. Market phenomenon in the past and now the development of the law of change and the level of development, will affect the market phenomenon in the future development of the law of change and the level of scale; market phenomenon in the future of the law of change and the level of change, is the result of the market phenomenon in the past and now the law of change and the level of development.

It should be pointed out that, as the development of things not only has the characteristics of continuity, but also complex and diverse. Therefore, in the application of time series analysis for market forecasting should pay attention to the market phenomenon of the future development of the law of change and the level of development, not necessarily with its history and the current development of the law of change is completely consistent. With the development of the market phenomenon, it will also appear some new features. Therefore, in the time series analysis and forecasting, must not be mechanically according to the market phenomenon in the past and the present law to extend outward. Must study and analyze the new characteristics of the market phenomenon changes, new performance, and these new features and new performance will be fully considered in the forecast value. So that the market phenomenon to make both the continuation of its historical pattern of change, but also in line with the reality of the performance of the reliable forecast results.

②Time series analysis forecasting method highlights the role of the time factor in the forecast, without taking into account the impact of specific external factors. Time series in the time series analysis forecasting method is at the core of the position, no time series, there is no existence of this method. Although, the development and change of the forecast object is affected by many factors. However, the use of time series analysis for quantitative forecasting, in fact, all the influencing factors attributed to the factor of time, only to recognize the combined effect of all the influencing factors, and in the future still play a role in the forecasting object, and did not go to analyze and explore the causal relationship between the forecasting object and the influencing factors. Therefore, in order to find accurate forecasts that reflect the future development of the market, the use of time series analysis for forecasting, it is necessary to combine quantitative and qualitative methods of analysis, from the qualitative aspect of the full study of the relationship between the various factors and the market, in the full analysis of the study of various factors affecting the market to determine the value of forecasts based on the full analysis of the various factors affecting the market changes.

It should be pointed out that the time series forecasting method for highlighting the time series does not take into account the impact of external factors, and thus there is a forecasting error defects, when encountered in the outside world, there are often large changes, often have a large deviation, time series forecasting method for short- and medium-term forecasts than the long-term forecasting effect is good. Because objective things, especially economic phenomena, in a longer period of time the possibility of changes in external factors, they must have a significant impact on the market economic phenomena. If this is the case, and forecasts are made only on the basis of the time factor, without taking into account the impact of external factors on the object of forecasting, the results of the forecasts will be seriously out of line with the actual state of affairs.

### Advantages and disadvantages of arima model

Introduction of ARIMA model

1, the so-called ARIMA model, refers to the non-stationary time series into a stationary time series, and then the dependent variable only to its lagged value as well as the present and lagged value of the random error term of the regression of the model established.

2, ARIMA model is called autoregressive moving average model, the full name is (ARIMA, AutoregressiveIntegratedMovingAverageModel). Also known as ARIMA (p, d, q), is a statistical model (statisticmodel) in the most common type of model used for time series forecasting.

3, AR, MA, ARMA are applied to the original data is a smooth time series. ARIMA is applied to the original data after the difference is a smooth time series. The time series are different AR (autoregressive model), AR (p), p-order autoregressive model. MA (moving average model), MA (q), q-order moving average model.

4, ARIMA model of the basic idea is: the forecast object over time as a random sequence of data, with a certain mathematical model to approximate the description of this sequence. Once this model is recognized it is possible to predict future values from the past and present values of the time series.

Data Analysis Techniques:AR/MA/ARMA/ARIMA Modeling System for Time Series Analysis

1. Because of the shortcomings of the traditional time series analysis techniques (time series decomposition), statisticians have developed more generalized methods for time series analysis, of which AR/MA/ARMA/ARIMA have played a very important role in this development process. role in this development process, until now, they all play an important role in the actual working life.

2. A time series is a set of data measured at consecutive times, which is mathematically defined as a set of vectors x(t), t=0, 1, 2, 3, … , where t denotes the point in time at which the data are located, and x(t) is a set of random variables arranged in chronological (measured) order.

3. The ARIMA model is for modeling non-stationary time series. In other words, non-stationary time series to build ARMA model, first need to be transformed into a smooth time series after the difference, and then build ARMA model.The principle of ARIMA model. As introduced earlier, ARIMA model is actually a combination of AR model and MA model.

4, the use of different objects AR, MA, ARMA are used in the original data is a smooth time series. ARIMA is used in the original data after the difference is a smooth time series. Time series different AR (autoregressive model), AR (p), p-order autoregressive model.

5. Clearly, the ARMA model describes a time-invariant linear system.? ARMA process with AR order p and MA order Q is often written as ARMA (p, q). ARIMA model, differential autoregressive sliding average model (sliding is also translated as moving), also known as summation autoregressive sliding average model, one of the methods of predictive analysis of time series.

Principal component regression model can forecast with the time series ARIMA forecasting model is also used to forecast, they…

Definition of time series Definition 1: A time series is a set of statistical data, according to the sequence of the time of its occurrence into a sequence. Definition 2: A sequence of successive observations of the same phenomenon at different times is called a time series.

It is often used to predict future values of time series data, such as stock prices, climate change, etc. Time series forecasting usually uses statistical methods to model time series, such as ARIMA (autoregressive moving average model) and ETS (exponential smoothing model).

ARIMA model is known as differential autoregressive moving average model: ARIMA model is a famous time series forecasting method proposed by Box and Jenkins in the early 70s, so it is also known as box-jenkins model, Box-Jenkins method.

### Characteristics and advantages and disadvantages of the smoothed exponential method?

Characteristics of the smoothing exponential method: the simple full-period averaging method is to make equal use of all the past data of the time series without omitting any of them; the moving average method disregards the more distant data and gives more weight to the recent information in the weighted moving average method; the exponential smoothing method is compatible with the full-period averaging and moving averaging, does not discard the past data, but only gives a gradually decreasing degree of influence, i.e., gives weights that gradually converge to zero as the data move away from the data. The exponential smoothing method is compatible with the full period average and the moving average, and does not discard past data, but gives only a diminishing degree of influence, i.e., weights that converge to zero as the data move away.

The advantages and disadvantages of the smoothing index method:

1. Advantages: less data required, you can predict the desired results, exponential smoothing method is based on the development of a moving average method of time series analysis and forecasting method, compatible with the full average and the moving average of the length of the period, do not give up the data in the past, but only to give gradually weakening the influence of the degree of the data away from, give gradually converge to zero weights. The weights that are given to gradually converge to zero are calculated by exponential smoothing, and the future of the phenomenon is predicted with a certain time series forecasting model.

2. Disadvantages: gives a smaller weight to the distant future and a larger weight to the near future, so only short-term forecasting.