Multiple regression analysis Paragonimus metacercariae excystation factors related to abnormal points of.doc

 PAGE \* MERGEFORMAT 3 Multiple regression analysis Paragonimus metacercariae excystation factors related to abnormal points of Author: Qu Gui Chen Jingwu Du defense [Abstract] Objective: To evaluate the accuracy found to deal reasonably with the experimental data outliers in the data to meet the multiple linear regression analysis of conditions to ensure authenticity and reliability of the results. Methods: The residual map, residual analysis, impact analysis, robust regression and in accordance with our professional expertise. Results: The first set of data points corresponding to outliers by deleting the outliers, the least square method to establish the equation and the establishment of robust regression line equation. Conclusion: The combination of several methods to determine outliers, and further enhance the professional conclusions of the scientific. [Keywords:] abnormal points; multiple regression analysis; robust regression In order to study lung fluke metacercariae excystation of the influencing factors, we conducted tests and their test data orthogonal multiple linear regression analysis, aimed at establishing a multiple linear regression model to analyze the influencing factors to explore the role of the nature and size. In the regression model, the exception is used to indicate a great deviation from the established model of data points. In particular, the data to focus on the statistics of those values have a very large influence on the impact of the strong points [1]. Once the abnormal data should be handled accordingly in a timely manner in order to improve the quality of the regression equation. 1 Materials and Methods 1  1 Data based on professional knowledge and influence the selection of Paragonimus metacercariae excystation four major factors and level of selected L16 (45) orthogonal table to be tested, test data collection, such as Table 1. Table 1 Orthogonal experiment data (omitted) 1  2 Studentized residuals in order to s