Compositional data, such as relative information, is a crucial aspect of machine learning and other related fields. It is typically recorded as closed data or sums to a constant, like 100%. The statistical linear mode...Compositional data, such as relative information, is a crucial aspect of machine learning and other related fields. It is typically recorded as closed data or sums to a constant, like 100%. The statistical linear model is the most used technique for identifying hidden relationships between underlying random variables of interest. However, data quality is a significant challenge in machine learning, especially when missing data is present. The linear regression model is a commonly used statistical modeling technique used in various applications to find relationships between variables of interest. When estimating linear regression parameters which are useful for things like future prediction and partial effects analysis of independent variables, maximum likelihood estimation (MLE) is the method of choice. However, many datasets contain missing observations, which can lead to costly and time-consuming data recovery. To address this issue, the expectation-maximization (EM) algorithm has been suggested as a solution for situations including missing data. The EM algorithm repeatedly finds the best estimates of parameters in statistical models that depend on variables or data that have not been observed. This is called maximum likelihood or maximum a posteriori (MAP). Using the present estimate as input, the expectation (E) step constructs a log-likelihood function. Finding the parameters that maximize the anticipated log-likelihood, as determined in the E step, is the job of the maximization (M) phase. This study looked at how well the EM algorithm worked on a made-up compositional dataset with missing observations. It used both the robust least square version and ordinary least square regression techniques. The efficacy of the EM algorithm was compared with two alternative imputation techniques, k-Nearest Neighbor (k-NN) and mean imputation (), in terms of Aitchison distances and covariance.展开更多
不平衡数据是监督学习中的一个挑战性问题。传统的分类器通常偏向多数类,忽略了少数类,而少数类样本往往包含很多重要信息,需要得到更多的关注。针对此问题,提出了一种基于密度峰值聚类算法的过采样技术(An Oversampling Technique base...不平衡数据是监督学习中的一个挑战性问题。传统的分类器通常偏向多数类,忽略了少数类,而少数类样本往往包含很多重要信息,需要得到更多的关注。针对此问题,提出了一种基于密度峰值聚类算法的过采样技术(An Oversampling Technique based on Density Peak Clustering,DPCOTE)。DPCOTE的主要思想是:(1)利用k近邻算法去除多数类和少数类噪声样本;(2)基于密度峰值聚类算法(Density peaks clustering algorithm,DPC)中的2个重要因子,即样本局部密度和样本到局部密度较高的最近邻的距离,来为每个少数类样本分配采样权重;(3)对于DPC算法中涉及到的距离,使用马氏距离来度量,以消除样本特征量纲不一致问题。最后,在12个UCI数据集上进行了对比实验,用不同的指标评价分类结果,结果表明本文提出的算法在处理不平衡分类问题时优于其它过采样方法。展开更多
Text categorization(TC)is one of the widely studied branches of text mining and has many applications in different domains.It tries to automatically assign a text document to one of the predefined categories often by ...Text categorization(TC)is one of the widely studied branches of text mining and has many applications in different domains.It tries to automatically assign a text document to one of the predefined categories often by using machine learning(ML)techniques.Choosing the best classifier in this task is the most important step in which k-Nearest Neighbor(KNN)is widely employed as a classifier as well as several other well-known ones such as Support Vector Machine,Multinomial Naive Bayes,Logistic Regression,and so on.The KNN has been extensively used for TC tasks and is one of the oldest and simplest methods for pattern classification.Its performance crucially relies on the distance metric used to identify nearest neighbors such that the most frequently observed label among these neighbors is used to classify an unseen test instance.Hence,in this paper,a comparative analysis of the KNN classifier is performed on a subset(i.e.,R8)of the Reuters-21578 benchmark dataset for TC.Experimental results are obtained by using different distance metrics as well as recently proposed distance learning metrics under different cases where the feature model and term weighting scheme are different.Our comparative evaluation of the results shows that Bray-Curtis and Linear Discriminant Analysis(LDA)are often superior to the other metrics and work well with raw term frequency weights.展开更多
文摘Compositional data, such as relative information, is a crucial aspect of machine learning and other related fields. It is typically recorded as closed data or sums to a constant, like 100%. The statistical linear model is the most used technique for identifying hidden relationships between underlying random variables of interest. However, data quality is a significant challenge in machine learning, especially when missing data is present. The linear regression model is a commonly used statistical modeling technique used in various applications to find relationships between variables of interest. When estimating linear regression parameters which are useful for things like future prediction and partial effects analysis of independent variables, maximum likelihood estimation (MLE) is the method of choice. However, many datasets contain missing observations, which can lead to costly and time-consuming data recovery. To address this issue, the expectation-maximization (EM) algorithm has been suggested as a solution for situations including missing data. The EM algorithm repeatedly finds the best estimates of parameters in statistical models that depend on variables or data that have not been observed. This is called maximum likelihood or maximum a posteriori (MAP). Using the present estimate as input, the expectation (E) step constructs a log-likelihood function. Finding the parameters that maximize the anticipated log-likelihood, as determined in the E step, is the job of the maximization (M) phase. This study looked at how well the EM algorithm worked on a made-up compositional dataset with missing observations. It used both the robust least square version and ordinary least square regression techniques. The efficacy of the EM algorithm was compared with two alternative imputation techniques, k-Nearest Neighbor (k-NN) and mean imputation (), in terms of Aitchison distances and covariance.
文摘不平衡数据是监督学习中的一个挑战性问题。传统的分类器通常偏向多数类,忽略了少数类,而少数类样本往往包含很多重要信息,需要得到更多的关注。针对此问题,提出了一种基于密度峰值聚类算法的过采样技术(An Oversampling Technique based on Density Peak Clustering,DPCOTE)。DPCOTE的主要思想是:(1)利用k近邻算法去除多数类和少数类噪声样本;(2)基于密度峰值聚类算法(Density peaks clustering algorithm,DPC)中的2个重要因子,即样本局部密度和样本到局部密度较高的最近邻的距离,来为每个少数类样本分配采样权重;(3)对于DPC算法中涉及到的距离,使用马氏距离来度量,以消除样本特征量纲不一致问题。最后,在12个UCI数据集上进行了对比实验,用不同的指标评价分类结果,结果表明本文提出的算法在处理不平衡分类问题时优于其它过采样方法。
文摘Text categorization(TC)is one of the widely studied branches of text mining and has many applications in different domains.It tries to automatically assign a text document to one of the predefined categories often by using machine learning(ML)techniques.Choosing the best classifier in this task is the most important step in which k-Nearest Neighbor(KNN)is widely employed as a classifier as well as several other well-known ones such as Support Vector Machine,Multinomial Naive Bayes,Logistic Regression,and so on.The KNN has been extensively used for TC tasks and is one of the oldest and simplest methods for pattern classification.Its performance crucially relies on the distance metric used to identify nearest neighbors such that the most frequently observed label among these neighbors is used to classify an unseen test instance.Hence,in this paper,a comparative analysis of the KNN classifier is performed on a subset(i.e.,R8)of the Reuters-21578 benchmark dataset for TC.Experimental results are obtained by using different distance metrics as well as recently proposed distance learning metrics under different cases where the feature model and term weighting scheme are different.Our comparative evaluation of the results shows that Bray-Curtis and Linear Discriminant Analysis(LDA)are often superior to the other metrics and work well with raw term frequency weights.
