朴素贝叶斯分类器是机器学习领域中一种重要的分类算法,根据该算法的前提,利用Foley-Sammon变换算法进行特征提取,提出了一种基于Foley-Sammon变换的朴素贝叶斯分类器NBFST(Na ve Bayesian Classifier with Foley-Sammon Transform).结...朴素贝叶斯分类器是机器学习领域中一种重要的分类算法,根据该算法的前提,利用Foley-Sammon变换算法进行特征提取,提出了一种基于Foley-Sammon变换的朴素贝叶斯分类器NBFST(Na ve Bayesian Classifier with Foley-Sammon Transform).结果表明,NBFST能够在大多数数据集上具有较高的分类准确率.展开更多
本文在仔细分析特征选择思想的基础上,将特征选择过程嵌入到学习机里面,提出了一种基于改进支持向量机的特征选择算法(Feature selection via Modified Support Vector Machines),该方法通过对特征的权重进行排序来实现特征选择.利用可...本文在仔细分析特征选择思想的基础上,将特征选择过程嵌入到学习机里面,提出了一种基于改进支持向量机的特征选择算法(Feature selection via Modified Support Vector Machines),该方法通过对特征的权重进行排序来实现特征选择.利用可以将特征选择过程和学习过程有机地统一起来,实验表明,与其它方法比较,该方法能够达到比较好的效果.展开更多
A modified genetic algorithm of multiple selection strategies, crossover strategies and adaptive operator is constructed, and it is used to estimate the kinetic parameters in autocatalytic oxidation of cyclohexane. Th...A modified genetic algorithm of multiple selection strategies, crossover strategies and adaptive operator is constructed, and it is used to estimate the kinetic parameters in autocatalytic oxidation of cyclohexane. The influences of selection strategy, crossover strategy and mutation strategy on algorithm performance are discussed. This algorithm with a specially designed adaptive operator avoids the problem of local optimum usually associated with using standard genetic algorithm and simplex method. The kinetic parameters obtained from the modified genetic algorithm are credible and the calculation results using these parameters agree well with experimental data. Furthermore, a new kinetic model of cyclohexane autocatalytic oxidation is established and the kinetic parameters are estimated by using the modified genetic algorithm.展开更多
文摘朴素贝叶斯分类器是机器学习领域中一种重要的分类算法,根据该算法的前提,利用Foley-Sammon变换算法进行特征提取,提出了一种基于Foley-Sammon变换的朴素贝叶斯分类器NBFST(Na ve Bayesian Classifier with Foley-Sammon Transform).结果表明,NBFST能够在大多数数据集上具有较高的分类准确率.
文摘本文在仔细分析特征选择思想的基础上,将特征选择过程嵌入到学习机里面,提出了一种基于改进支持向量机的特征选择算法(Feature selection via Modified Support Vector Machines),该方法通过对特征的权重进行排序来实现特征选择.利用可以将特征选择过程和学习过程有机地统一起来,实验表明,与其它方法比较,该方法能够达到比较好的效果.
基金Supported by the National Natural Science Foundation of China (No. 20176046).
文摘A modified genetic algorithm of multiple selection strategies, crossover strategies and adaptive operator is constructed, and it is used to estimate the kinetic parameters in autocatalytic oxidation of cyclohexane. The influences of selection strategy, crossover strategy and mutation strategy on algorithm performance are discussed. This algorithm with a specially designed adaptive operator avoids the problem of local optimum usually associated with using standard genetic algorithm and simplex method. The kinetic parameters obtained from the modified genetic algorithm are credible and the calculation results using these parameters agree well with experimental data. Furthermore, a new kinetic model of cyclohexane autocatalytic oxidation is established and the kinetic parameters are estimated by using the modified genetic algorithm.
