Different choice of appropriate arma model using the information criterion: Applying the share certificate of silicone valley companies

In many cases, it is often difficult to identify dependent variables or independent variables that will affect variables when Regression Analysis is applied. A large number of problems in the statistical inference may be related to statistical modeling problems. Model Selection is the process of choosing the most suitable model from the existing candidate models by using the simplest model, revealing which of the explanatory variables or which ones are effective on the explained variable. Choosing the most appropriate model with the simplest model has both computational comfort during analysis and mitigating effect on the cost table in the data compilation. By reducing the size of the problem, it is also possible to estimate the model parameters more accurately and make clearer determinations. In this study, the values of the daily stocks of the ten largest companies in the Silicon Valley during the one month period and the AIC, AICc and BIC information criteria in the ARMA model time series analysis were used to determine the appropriate variables for the optimal model. Key Words: Model Selection, Information Criteria, Silicon Valley, Technology, ARMA Models, AIC, BIC, AICc