The analysis of Turkish import with Ridge regression

Ridge regression and principal component regression as biased estimation methods have been applied to data at hand as well as time series methodology. To confirm the reliability of all the results of obtained by all methodologies mentioned above, a comparison between all application results have been made. First of all, a model has been reached from the data with least of square. During the application the multicollinearity problem has been figured out. Since the data has a multicollinearity problem, the ridge regression analysis has been applied. Its results and least square model has been compared to each other. After that, for the same data, time series analysis and principal component regression analysis have been executed. Both the results of ridge regression and principal componment regression analysis have had an parallelity. Then the results of ridge regression and other analysis have been compared to each other. Finaly it has seen that the results of ridge regression supports the economical theory. At the end of the all analysis and comprasions; it has been revealed that the relative price doesn?t affect the import too much. Excel 2003, SPSS 11.5, Eviews 5 and NCSS 2007 have been implied for these applications. As a results of ridge regression analysis which has been applied for handled data set, it has been seen that import isn?t affected by the relative price. At the same time has been interpreted that these results can be reliable. Key words: Ridge regression, multicollinearity, least of square (LS), linear regression, nonlinear regression , time series, principal component regression