The feature-selection problem in training AdaBoost classifiers is addressed in this paper. A working feature subset is generated by adopting a novel feature subset selection method based on the partial least square （...The feature-selection problem in training AdaBoost classifiers is addressed in this paper. A working feature subset is generated by adopting a novel feature subset selection method based on the partial least square （PLS） regression, and then trained and selected from this feature subset in Boosting. The experiments show that the proposed PLS-based feature-selection method outperforms the current feature ranking method and the random sampling method.展开更多
In this paper, we provide an aggregate function homotopy interior point method to solve a class of Brouwer fixed-point problems. Compared with the homotopy method （proposed by Yu and Lin, Appl. Math. Comput., 74（199...In this paper, we provide an aggregate function homotopy interior point method to solve a class of Brouwer fixed-point problems. Compared with the homotopy method （proposed by Yu and Lin, Appl. Math. Comput., 74（1996）, 65）, the main adavantages of this method are as foUows： on the one hand, it can solve the Brouwer fixed-point problems in a broader class of nonconvex subsets Ω in R^n （in this paper, we let Ω={x∈ R^n ： gi（x） ≤0, i= 1,... , m}）; on the other hand, it can also deal with the subsets Ω with larger amount of constraints more effectively.展开更多
In this paper, we modify the homotopy method （proposed by Yu and Lin, Appl. Math. Comput., 74（1996）, 65） and hence make the modified method be able to solve Brouwer fixed-point problems in a broader class of nonco...In this paper, we modify the homotopy method （proposed by Yu and Lin, Appl. Math. Comput., 74（1996）, 65） and hence make the modified method be able to solve Brouwer fixed-point problems in a broader class of nonconvex subsets in Rn. In addition, a simple example is given to show the effectiveness of the modified method.展开更多
Many medical diagnosis applications are characterized by datasets that contain under-represented classes due to the fact that the disease is much rarer than the normal case. In such a situation classifiers such as dec...Many medical diagnosis applications are characterized by datasets that contain under-represented classes due to the fact that the disease is much rarer than the normal case. In such a situation classifiers such as decision trees and Na?ve Bayesian that generalize over the data are not the proper choice as classification methods. Case-based classifiers that can work on the samples seen so far are more appropriate for such a task. We propose to calculate the contingency table and class specific evaluation measures despite the overall accuracy for evaluation purposes of classifiers for these specific data characteristics. We evaluate the different options of our case-based classifier and compare the perform-ance to decision trees and Na?ve Bayesian. Finally, we give an outlook for further work.展开更多
This article compares the size of selected subsets using nonparametric subset selection rules with two different scoring rules for the observations. The scoring rules are based on the expected values of order statisti...This article compares the size of selected subsets using nonparametric subset selection rules with two different scoring rules for the observations. The scoring rules are based on the expected values of order statistics of the uniform distribution (yielding rank values) and of the normal distribution (yielding normal score values). The comparison is made using state motor vehicle traffic fatality rates, published in a 2016 article, with fifty-one states (including DC as a state) and over a nineteen-year period (1994 through 2012). The earlier study considered four block design selection rules—two for choosing a subset to contain the “best” population (i.e., state with lowest mean fatality rate) and two for the “worst” population (i.e., highest mean rate) with a probability of correct selection chosen to be 0.90. Two selection rules based on normal scores resulted in selected subset sizes substantially smaller than corresponding rules based on ranks (7 vs. 16 and 3 vs. 12). For two other selection rules, the subsets chosen were very close in size (within one). A comparison is also made using state homicide rates, published in a 2022 article, with fifty states and covering eight years. The results are qualitatively the same as those obtained with the motor vehicle traffic fatality rates.展开更多
The fatigue of heavy-haul railway bridges is considered a key concern due to high stress levels and cyclic loading.The evaluation of fatigue reliability is required to include factor correlations.A major challenge is ...The fatigue of heavy-haul railway bridges is considered a key concern due to high stress levels and cyclic loading.The evaluation of fatigue reliability is required to include factor correlations.A major challenge is presented by the construction of the cumulative distribution function(CDF)and the description of correlations between random variables.In this study,the copula function is used to analyze the fatigue failure probability of the Shuohuang heavy-haul railway bridge.A C-vine copula(CVC)-based joint probability density function(JPDF)is derived with eight correlated parameters.To enhance efficiency in small failure probability calculations,the subset simulation and most probable point(MPP)Monte Carlo importance sampling are introduced based on the Rosenblatt transform and C-vine model.Comparisons with traditional Monte Carlo methods confirm that high accuracy and efficiency are achieved.The results show that when parameter correlations are ignored,failure probability is underestimated,increasing safety risks in bridge assessments.展开更多
基金Supported by the National Natural Science Foundation of China(60772066)
文摘The feature-selection problem in training AdaBoost classifiers is addressed in this paper. A working feature subset is generated by adopting a novel feature subset selection method based on the partial least square （PLS） regression, and then trained and selected from this feature subset in Boosting. The experiments show that the proposed PLS-based feature-selection method outperforms the current feature ranking method and the random sampling method.
文摘In this paper, we provide an aggregate function homotopy interior point method to solve a class of Brouwer fixed-point problems. Compared with the homotopy method （proposed by Yu and Lin, Appl. Math. Comput., 74（1996）, 65）, the main adavantages of this method are as foUows： on the one hand, it can solve the Brouwer fixed-point problems in a broader class of nonconvex subsets Ω in R^n （in this paper, we let Ω={x∈ R^n ： gi（x） ≤0, i= 1,... , m}）; on the other hand, it can also deal with the subsets Ω with larger amount of constraints more effectively.
文摘In this paper, we modify the homotopy method （proposed by Yu and Lin, Appl. Math. Comput., 74（1996）, 65） and hence make the modified method be able to solve Brouwer fixed-point problems in a broader class of nonconvex subsets in Rn. In addition, a simple example is given to show the effectiveness of the modified method.
文摘Many medical diagnosis applications are characterized by datasets that contain under-represented classes due to the fact that the disease is much rarer than the normal case. In such a situation classifiers such as decision trees and Na?ve Bayesian that generalize over the data are not the proper choice as classification methods. Case-based classifiers that can work on the samples seen so far are more appropriate for such a task. We propose to calculate the contingency table and class specific evaluation measures despite the overall accuracy for evaluation purposes of classifiers for these specific data characteristics. We evaluate the different options of our case-based classifier and compare the perform-ance to decision trees and Na?ve Bayesian. Finally, we give an outlook for further work.
文摘This article compares the size of selected subsets using nonparametric subset selection rules with two different scoring rules for the observations. The scoring rules are based on the expected values of order statistics of the uniform distribution (yielding rank values) and of the normal distribution (yielding normal score values). The comparison is made using state motor vehicle traffic fatality rates, published in a 2016 article, with fifty-one states (including DC as a state) and over a nineteen-year period (1994 through 2012). The earlier study considered four block design selection rules—two for choosing a subset to contain the “best” population (i.e., state with lowest mean fatality rate) and two for the “worst” population (i.e., highest mean rate) with a probability of correct selection chosen to be 0.90. Two selection rules based on normal scores resulted in selected subset sizes substantially smaller than corresponding rules based on ranks (7 vs. 16 and 3 vs. 12). For two other selection rules, the subsets chosen were very close in size (within one). A comparison is also made using state homicide rates, published in a 2022 article, with fifty states and covering eight years. The results are qualitatively the same as those obtained with the motor vehicle traffic fatality rates.
基金Project supported by the National Natural Science Foundation of China(No.52278180)。
文摘The fatigue of heavy-haul railway bridges is considered a key concern due to high stress levels and cyclic loading.The evaluation of fatigue reliability is required to include factor correlations.A major challenge is presented by the construction of the cumulative distribution function(CDF)and the description of correlations between random variables.In this study,the copula function is used to analyze the fatigue failure probability of the Shuohuang heavy-haul railway bridge.A C-vine copula(CVC)-based joint probability density function(JPDF)is derived with eight correlated parameters.To enhance efficiency in small failure probability calculations,the subset simulation and most probable point(MPP)Monte Carlo importance sampling are introduced based on the Rosenblatt transform and C-vine model.Comparisons with traditional Monte Carlo methods confirm that high accuracy and efficiency are achieved.The results show that when parameter correlations are ignored,failure probability is underestimated,increasing safety risks in bridge assessments.
