Feature selection based on label distribution and fuzzy mutual information

Existing feature selection methods mainly target single-label learning and multi-label learning, and only a few algorithms are optimized for label distribution learning. In label distribution learning, the associated labels of each sample have different levels of importance. Therefore, multi-label feature selection algorithms cannot be directly applied to label distribution learning. Discretizing label distribution data into multi-label data will cause part of the supervision information to be lost. In most practical applications of label distribution learning, the feature space is undefined, and the features are in the form of flow features. To solve this problem, this paper applies fuzzy rough set theory and applies the flow feature framework to propose a dynamic label distribution feature selection algorithm that handles flow features. Experimental results show that the proposed method is more effective than six state-of-the-art feature selection algorithms on 12 datasets with respect to six representative evaluation metrics.

As we all know, the multi-label learning is faced with the "dimension disaster", and the label distribution learning is actually faced with the same problem. Therefore, it is necessary to perform some appropriate pre-processing on the data before the label distribution learning, such as feature selection. Nowadays, researchers have proposed many feature selection algorithms for multi-label learning, while feature selection algorithms for label distribution learning are few. The difference between label distribution learning and multi-label learning is that in the existing multi-label classification methods, the importance of each label is considered to be the same, but in actual cases, different labels has different importance when they are describing the same instance, that is, there is an imbalance between the labels. Therefore, this paper proposed a feature selection algorithm based on information measures in fuzzy rough set theory for label distribution learning. The mutual information in fuzzy rough set theory is used to calculate the correlation between features and labels, and then a feature subset is selected by correlation maximization and redundancy minimization strategy. Finally, the experimental results demonstrate that the proposed feature selection algorithm is able to select valid subsets of features.

Robust feature selection using label enhancement and β-precision fuzzy rough sets for multilabel fuzzy decision system

Text datasets have many terms, which decrease the classification accuracy. According to the high-dimensional text data, there are more challenges for these methods. Each classification method has strengths and weaknesses in its feature selection function. Therefore, ensembling should be used for better classifications and exploitation of strengths. In this paper, for the first time, we have presented an ensemble multi-label (ML) feature selection method for the text datasets using the Geometric-Mean (GM) aggregation approach. For this purpose, we have used four multi-label feature selection (MLFS) algorithms with different structures to achieve a good result. Then, the performance and results obtained by the GM method are compared with the four algorithms and based on the six classification criteria on three ML datasets with text domains. According to the obtained outputs, it is possible to realize the ability of the GMA (proposed method) method in using strengths and ignoring weaknesses in the path of feature selection, and be more accurate. Finally, according to experiments and obtained results, superiority of the GMA over other methods can be well seen.

A label-specific multi-label feature selection algorithm based on the Pareto dominance concept

Feature selection for multilabel classification with missing labels via multi-scale fusion fuzzy uncertainty measures

Multi-label feature selection, which is an efficient and effective pre-processing step in machine learning and data mining, can select a feature subset that contains more contributions for multi-label classification while improving the performance of the classifiers. In real-world applications, an instance may be associated with multiple related labels with different relative importances, and the process of obtaining different features usually requires different costs, containing money, and time, etc. However, most existing works with regard to multi-label feature selection do not take into consideration the above two critical issues simultaneously. Therefore, in this paper, we exploit the idea of neighborhood granularity to enhance the traditional logical labels into label distribution forms to excavate the deeper supervised information hidden in multi-label data, and further consider the effect of the test cost under three different distributions, simultaneously. Motivated by these issues, a novel test cost multi-label feature selection algorithm with label enhancement and neighborhood granularity is designed. Moreover, the proposed algorithm is tested upon ten publicly available benchmark multi-label datasets with six widely-used metrics from two different aspects. Finally, two groups of experimental results demonstrate that the proposed algorithm achieves the satisfactory and superior performance over other four state-of-the-art comparing algorithms, and it is effective for improving the learning performance and decreasing the total test costs of the selected feature subset.

Label distribution learning (LDL) can more accurately represent the degree of correlation between labels and samples than multi-label learning. However, LDL usually has limited available label data, which is not conducive to training deep learning models. Data augmentation refers to methods for solving limited data problems by introducing unobserved data or latent variables to increase the size and quality of the training dataset. In this paper, we augment the dataset by mapping the features and label distributions of the generated samples to the same subspace, and using the generator to learn the distribution of the original data in this space. First, we use the encoder and generator to extract effective information from sample features and label distributions, respectively. Second, we randomly fuse the existing label distribution to generate a new label distribution, then map it to the subspace and restore the corresponding features through the decoder. Finally, these generated samples are mixed with the original training set to train a model for predicting label distribution. Experimental results on nine real-world datasets show that our proposed algorithm can improve the performance of deep learning models to a certain extent.

Enhanced multi-label feature selection considering label-specific relevant information

Label Distribution Learning (LDL), as a more general learning setting than generic single-label and multi-label learning, has been commonly used in computer vision and many other applications. To date, existing LDL approaches are designed and applied to data without considering the interdependence between instances. In this paper, we propose a Graph Label Distribution Learning (GLDL) framework, which explicitly models three types of relationships: instance-instance, label-label, and instance-label, to learn the label distribution for networked data. A label-label network is learned to capture label-to-label correlation, through which GLDL can accurately learn label distributions for nodes. Dual graph convolution network (GCN) Co-training with heterogeneous message passing ensures two GCNs, one focusing on instance-instance relationship and the other one targeting label-label correlation, are jointly trained such that instance-instance relationship can help induce label-label correlation and vice versa. Our theoretical study derives the error bound of GLDL. For verification, four benchmark datasets with label distributions for nodes are created using common graph benchmarks. The experiments show that considering dependency helps learn better label distributions for networked data, compared to state-of-the-art LDL baseline. In addition, GLDL not only outperforms simple GCN and graph attention networks (GAT) using distribution loss but is also superior to its variant considering label-label relationship as a static network. GLDL and its benchmarks are the first research endeavors to address LDL for graphs. Code and benchmark data are released for public access.

Feature selection can remove non-important features from the data and promote better classifiers. This task, when applied to multi-label data where each instance is associated with a set of labels, supports emerging applications. Although multi-label data usually exhibit label relations, label dependence has been little studied in feature selection. We proposed two multi-label feature selection algorithms that consider label relations. These methods were experimentally competitive with traditional approaches. Moreover, this work conducted a systematic literature review, summarizing 74 related papers.

Due to the existence of text data, multi-label (ML) text classification is an essential task in machine learning. Feature selection is an essential and effective preprocess to enhance the learning process. Choosing a Multi-Label Feature Selection (MLFS) algorithm is the most basic, critical, and sensitive choice in ML classification operations. If this choice is based on a criterion, it cannot be attributed to always being sound. Choosing the best algorithm must be evaluated using several different criteria to be examined from different aspects. In this article, we turn the issue into an election and use the Weighted Borda Count method for voting. We do the voting in three stages continuously so that a subset of different features does the voting. In the second stage, voting of different methods is done with six criteria, and each criterion selects the methods in order of priority from the beginning to the end. Voting steps 1 and 2 are performed on eighteen text datasets used. Finally, in the final voting stage, the methods are evaluated and voted on by different text datasets. The final result of the voting in the third stage shows the desired MLFS methods based on their performance from beginning to end. According to the experiments performed and the results obtained, it can be seen that the selection of the algorithm based on several different criteria and considering the overall performance of the algorithm will be better than the selection based on one criterion.

In a conventional supervised learning paradigm, each data instance is associated with one single class label. Multilabel learning differs in the way that data instances may belong to multiple concepts simultaneously, which naturally appear in a variety of high impact domains, ranging from bioinformatics and information retrieval to multimedia analysis. It targets leveraging the multiple label information of data instances to build a predictive learning model that can classify unlabeled instances into one or multiple predefined target classes. In multilabel learning, even though each instance is associated with a rich set of class labels, the label information could be noisy and incomplete as the labeling process is both time consuming and labor expensive, leading to potential missing annotations or even erroneous annotations. The existence of noisy and missing labels could negatively affect the performance of underlying learning algorithms. More often than not, multilabeled data often has noisy, irrelevant, and redundant features of high dimensionality. The existence of these uninformative features may also deteriorate the predictive power of the learning model due to the curse of dimensionality. Feature selection, as an effective dimensionality reduction technique, has shown to be powerful in preparing high-dimensional data for numerous data mining and machine-learning tasks. However, a vast majority of existing multilabel feature selection algorithms either boil down to solving multiple single-labeled feature selection problems or directly make use of the imperfect labels to guide the selection of representative features. As a result, they may not be able to obtain discriminative features shared across multiple labels. In this article, to bridge the gap between a rich source of multilabel information and its blemish in practical usage, we propose a novel noise-resilient multilabel informed feature selection framework (MIFS) by exploiting the correlations among different labels. In particular, to reduce the negative effects of imperfect label information in obtaining label correlations, we decompose the multilabel information of data instances into a low-dimensional space and then employ the reduced label representation to guide the feature selection phase via a joint sparse regression framework. Empirical studies on both synthetic and real-world datasets demonstrate the effectiveness and efficiency of the proposed MIFS framework.

Recently, multilabel classification has received significant attention during the past years. A multilabel classification approach called coupled k-nearest neighbors algorithm for multilabel classification (called here as CK-STC) reported in the literature exploits coupled label similarities between the labels and provides improved performance [Liu and Cao in A Coupled k-Nearest Neighbor Algorithm for Multi-label Classification, pp. 176–187, 2015]. A multilabel feature selection is presented in Li et al. [Multi-label Feature Selection via Information Gain, pp. 346–355, 2014] and called as FSVIG here. FSVIG uses information gain that shows better performance when used with ML-NB, ML-kNN, and RandSvm when compared with existing multilabel feature selection algorithms.This paper investigates the performance of FSVIG when used with CK-STC and compares its performance with other multilabel feature selection algorithms available in MULAN using standard multilabel datasets. Experimental results show that FSVIG when used with CK-STC provides better performance in terms of average precision and one-error.

Multilabel learning (MLL), as a hot topic in the field of machine learning, has attracted wide attention from many scholars due to its ability to express output space polysemy. In recent years, a large number of achievements about MLL have emerged. Among these achievements, there are several typical issues worthy of attention. Firstly, the correlation among labels plays a key role in improving MLL model training process. Many MLL algorithms try to fully and effectively use the correlation among labels to improve the performance. Secondly, existing MLL evaluation metrics, which is different from those in binary classification, often reflects the generalization performance of MLL classifiers in some aspects. How to choose metrics in algorithms to improve their generalization performance and fairness is another issue that should be concerned. Thirdly, in many practical MLL applications, there are many unlabeled instances due to their labeling cost in training datasets. How to use the wealth information contained in the correlation among unlabeled instances may contribute to reducing of the labeling cost in MLL and improving performance. Fourthly, labels assigned to instances may not be equally descriptive in many applications. How to describe the importance of each label in output space to an instance has become one of research points that many scholars have paid attention to in recent years. This paper reviews the MLL-related research results of correlation among labels, evaluation metric, multilabel semisupervised learning, and label distribution learning (LDL) from a theoretical and algorithmic perspective. Finally, the related research work on MLL is summarized and discussed.