Papers by Keyword: Box-Cox Transformation

Paper TitlePage

Authors: Ya Wen Hou, Bin Hui Wang
Abstract: Assuming that a process is subject to the normal distribution when calculating the process capability index traditionally. A main flaw lies in the process capability index is very sensitive to such changes if normal hypothesis was not satisfied. In order to obtain accurate process capability index of quality characteristics under this circumstance, this paper adopts Rosenblatt transformation to compare with Box-Cox transformation and Johnson transformation with type 6110 Connecting Rod Bush radius as an example, the results show that Rosenblatt transformation proposed performs better and implement is simple.
Authors: Wen Hua Shi, Chun Liang Chen, Jin Tao Niu
Abstract: Abstract: Formerly, the research to the assembly process of gear was commonly based on the normal assumption. However, in practice the clearance between gears in mesh does not necessarily obey normal distribution. Based on the mentioned above, the non-normal process capability analysis is fulfilled with the Box-Cox transformation and the data collected in the workshop. The corresponding result is compared with the directly obtained result, which validates the rationality and effectiveness.
Authors: Hai Zhen Wen, Xiao Qing Bu, Ling Zhang
Abstract: Box-Cox transformation allows functional forms more flexible. On the basis of the principle of model optimization, an empirical study is made for housing market of Hangzhou City. By collecting 2417 housing data in Hangzhou City, a housing hedonic price model with Box-Cox transformations is set up with 18 factors as housing characteristics. The model is estimated after the grid-search procedure by using MATLAB and SPSS software, and the statistical test shows that the logarithmic function is the optimal form. The model comparisons in the fitness and forecasting performance indicate that the logarithmic model is superior to other three models of the linear, semi-logarithmic and inverse semi-logarithmic. Empirical analysis suggests that the Box-Cox transformation is valid and feasible in choosing functional forms, can be used to optimize hedonic price models.
Authors: Pi Yun Chen, Yu Yi Fu, Kuo Lan Su, Jin Tsong Jeng
Abstract: In this paper, the Box–Cox transformation-based annealing robust fuzzy neural networks (ARFNNs) are proposed for identification of the nonlinear Magneto-rheological (MR) damper with outliers and skewness noises. Firstly, utilizing the Box-Cox transformation that its object is usually to make residuals more homogeneous in regression, or transform data to be normally distributed. Consequently, a support vector regression (SVR) method with Gaussian kernel function has the good performance to determine the number of rule in the simplified fuzzy inference systems and initial weights in the fuzzy neural networks. Finally, the annealing robust learning algorithm (ARLA) can be used effectively to adjust the parameters of the Box-Cox transformation-based ARFNNs. Simulation results show the superiority of the proposed method for the nonlinear MR damper systems with outliers and skewness noises.
Showing 1 to 4 of 4 Paper Titles