Sparse feature selection: Relevance, redundancy and locality structure preserving guided by pairwise constraints

Feature selection is an important pre-processing stage in many machine learning and pattern recognition tasks, which eliminates irrelevant and redundant features and improves learning performance. Regularized sparse feature selection methods like Lasso and its variants using l 1 -norm regularization term in their optimization problem have received much attention in recent years. Prior information could be represented as the class labels or pairwise constraints, i.e., must-link (positive) and cannot-link (negative) constraints. In this paper, besides the l 1 -norm regularization term, a normalized adapted affinity matrix is applied to embed the pairwise constraints in the affinity matrix. In the proposed affinity matrix, the weights are strengthened/weakened according to the positive/negative constraints. The experimental results on several data sets from University of California-Irvine (UCI) machine learning repository and a high dimensional data set, show the effectiveness of the proposed method in the classification tasks compared to some similar feature selection methods.

LSFSR: Local label correlation-based sparse multilabel feature selection with feature redundancy

With the rapid development of artificial intelligence in recent years, the research on image processing, text mining, and genome informatics has gradually deepened, and the mining of large-scale databases has begun to receive more and more attention. The objects of data mining have also become more complex, and the data dimensions of mining objects have become higher and higher. Compared with the ultra-high data dimensions, the number of samples available for analysis is too small, resulting in the production of high-dimensional small sample data. High-dimensional small sample data will bring serious dimensional disasters to the mining process. Through feature selection, redundancy and noise features in high-dimensional small sample data can be effectively eliminated, avoiding dimensional disasters and improving the actual efficiency of mining algorithms. However, the existing feature selection methods emphasize the classification or clustering performance of the feature selection results and ignore the stability of the feature selection results, which will lead to unstable feature selection results, and it is difficult to obtain real and understandable features. Based on the traditional feature selection method, this paper proposes an ensemble feature selection method, Random Bits Forest Recursive Clustering Eliminate (RBF-RCE) feature selection method, combined with multiple sets of basic classifiers to carry out parallel learning and screen out the best feature classification results, optimizes the classification performance of traditional feature selection methods, and can also improve the stability of feature selection. Then, this paper analyzes the reasons for the instability of feature selection and introduces a feature selection stability measurement method, the Intersection Measurement (IM), to evaluate whether the feature selection process is stable. The effectiveness of the proposed method is verified by experiments on several groups of high-dimensional small sample data sets.

Background: Dimension disaster is often associated with feature extraction. The extracted features may contain more redundant feature information, which leads to the limitation of computing ability and overfitting problems. Objective: Feature selection is an important strategy to overcome the problems from dimension disaster. In most machine learning tasks, features determine the upper limit of the model performance. Therefore, more and more feature selection methods should be developed to optimize redundant features. Methods: In this paper, we introduce a new technique to optimize sequence features based on the binomial distribution (BD). Firstly, the principle of the binomial distribution algorithm is introduced in detail. Then, the proposed algorithm is compared with other commonly used feature selection methods on three different types of datasets by using a Random Forest classifier with the same parameters. Results: The results confirm that BD has a promising improvement in feature selection and classification accuracy. Conclusion: Finally, we provide the source code and executable program package (http: //lingroup. cn/server/BDselect/), by which users can easily perform our algorithm in their researches.

In many real-world unsupervised learning applications, given data with balanced distribution, that is, there are an approximately equal number of instances in each class, we often need to construct a model to reveal such balance. However, in many data, especially the high-dimensional ones, the data in the original feature space often do not present such balance due to the redundant and noisy features. To tackle this problem, we apply an unsupervised spectral feature selection method to select some informative features, which can better reveal the balanced structure of data. Although spectral feature selection is one of the most popular unsupervised feature selection methods and has been widely studied, none of the existing spectral feature selection methods consider the balance property of data. To address this issue, in this article, we propose a novel balanced spectral feature selection (BSFS) method, which not only selects the discriminative features but also picks those to reveal the balanced structure of data. To the best of our knowledge, this is the first spectral feature selection method considering balance structure of data. By introducing a balanced regularization term, we integrate the balanced spectral clustering and feature selection into a unified framework seamlessly. At last, the experiments on benchmark datasets show that the proposed one outperforms the conventional feature selection methods in both clustering performance and balance, which demonstrates the effectiveness and efficiency of the proposed method.

Software defect prediction models are used for predicting high risk software components. Feature selection has significant impact on the prediction performance of the software defect prediction models since redundant and unimportant features make the prediction model more difficult to learn. Ensemble feature selection has recently emerged as a new methodology for enhancing feature selection performance. This paper proposes a new multi-criteria-decision-making (MCDM) based ensemble feature selection (EFS) method. This new method is termed as MCDM-EFS. The proposed method, MCDM-EFS, first generates the decision matrix signifying the feature’s importance score with respect to various existing feature selection methods. Next, the decision matrix is used as the input to well-known MCDM method TOPSIS for assigning a final rank to each feature. The proposed approach is validated by an experimental study for predicting software defects using two classifiers K-nearest neighbor (KNN) and naïve bayes (NB) over five open-source datasets. The predictive performance of the proposed approach is compared with existing feature selection algorithms. Two evaluation metrics – nMCC and G-measure are used to compare predictive performance. The experimental results show that the MCDM-EFS significantly improves the predictive performance of software defect prediction models against other feature selection methods in terms of nMCC as well as G-measure.

Graph self-representation method for unsupervised feature selection

Efficient nonconvex sparse group feature selection via continuous and discrete optimization

Many feature selection methods based on supervised learning have investigated side information that considers labelled instances. In this paper, we present a new feature selection method for semi-supervised learning, helped by side information, and the selected features are useful for discovering natural clusters. To select salient features for clustering, most existing wrapper-based feature selection methods study features from the partitioned clusters that result from a prespecified clustering algorithm. The selected features often suffer from both noisy features participating in the partitioning process and bias in the clustering algorithm selection. To overcome these problems, our method does not need clustered information to train features. In particular, our method uses side information to select features directly from the nearest and the farthest neighbours regardless of any clustering algorithm helping to partition data into clusters; specifically, the side information is considered to form pairwise instance constraints to guide the feature selection process and assign a weight for each instance for training. Experimental results indicate that our method, tested on image datasets, outperforms many well known wrapper-based feature selection methods in selecting features for clustering.

In the past decade, big data has become increasingly prevalent in a large number of applications. As a result, datasets suffering from noise and redundancy issues have necessitated the use of feature selection across multiple domains. However, a common concern in feature selection is that different approaches can give very different results when applied to similar datasets. Aggregating the results of different selection methods helps to resolve this concern and control the diversity of selected feature subsets. In this work, we implemented a general framework for the ensemble of multiple feature selection methods. Based on diversified datasets generated from the original set of observations, we aggregated the importance scores generated by multiple feature selection techniques using two methods: the Within Aggregation Method (WAM), which refers to aggregating importance scores within a single feature selection; and the Between Aggregation Method (BAM), which refers to aggregating importance scores between multiple feature selection methods. We applied the proposed framework on 13 real datasets with diverse performances and characteristics. The experimental evaluation showed that WAM provides an effective tool for determining the best feature selection method for a given dataset. WAM has also shown greater stability than BAM in terms of identifying important features. The computational demands of the two methods appeared to be comparable. The results of this work suggest that by applying both WAM and BAM, practitioners can gain a deeper understanding of the feature selection process.

In the recent years, high-dimensional data visualization has become a challenging task in data science and machine learning. As one of the most effective methods for high-dimensional data visualization, Parallel Coordinates Plots (PCPs) demonstrate dimensional reduction by transforming features of multivariate data into 2D axes. Such approach, however, does not consider the irrelevant or redundant features such that each feature is projected into the axis in a fixed manner. This paper proposed a novel PCP introduced by an unsupervised feature selection called Laplacian Score, which can be used to improve the visualization performance of PCP by ranking the importance of attributes based on their locality preserving power. The experimental results demonstrated that the performance of PCP visualization can be improved by feature selection method. Furthermore, we proposed a flexible user interface based on PCP visualization and Laplacian Score.

Ridge Regression (RR) is a classical method that is widely used in multiple regression analysis. However, traditional RR does not take the local geometric structure of data into consideration for discriminative learning and it is sensitive to outliers as it is based on <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{2}$</tex-math></inline-formula> -norm. To address this problem, this article proposes a novel method called Joint Sparse Locality Preserving Regression (JSLPR) for discriminative learning. JSLPR not only applies <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{2,1}$</tex-math></inline-formula> -norm on both loss function and regularization term but also takes the local geometric structure of the data into consideration. The use of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$L_{2,1}$</tex-math></inline-formula> -norm can guarantee the robustness to outliers or noises and the joint sparsity for effective feature selection. Taking the local geometric structure into consideration can improve the performance of the feature extraction and selection method when the data lie on a manifold. To solve the optimization problem of JSLPR, an iterative algorithm is proposed and the convergence of the algorithm is also proven. Experiments on four famous face databases are conducted and the results show the merit of the proposed JSLPR on feature extraction and selection.

A Bayesian classifier for sparsity-promoting feature selection is developed in this paper, where a set of nonlinear mappings for the original data is performed as a pre-processing step. The linear classification model with such mappings from the original input space to a nonlinear transformation space can not only construct the nonlinear classification boundary, but also realize the feature selection for the original data. A zero-mean Gaussian prior with Gamma precision and a finite approximation of Beta process prior are used to promote sparsity in the utilization of features and nonlinear mappings in our model, respectively. We derive the Variational Bayesian (VB) inference algorithm for the proposed linear classifier. Experimental results based on the synthetic data set, measured radar data set, high-dimensional gene expression data set, and several benchmark data sets demonstrate the aggressive and robust feature selection capability and comparable classification accuracy of our method comparing with some other existing classifiers.

QSAR analysis of diaryl COX-2 inhibitors: Comparison of feature selection and train-test data selection methods

The technological progress has lead to the generation and storage of abundant amounts of data. The extraction of information from such data has required the formulation of new data analysis tools. In this context, the Knowledge Discovery from Databases process was introduced. It is focused on the identification of valid, new, potentially useful, and comprehensible patterns in large databases. In this process, the task of finding patterns in data is usually called Data Mining. The efficacy and efficiency of data mining algorithms are directly influenced by the amount and quality of the data being analyzed. Redundant and/or uninformative features may make the data mining process inefficient. In this context, feature selection methods that can remove such features are frequently used. This work proposes a feature selection algorithm and some of its variants that are capable of identifying redundant features through clustering. The identification of redundant features can favor not only the pattern recognition process but also the comprehensibility of the obtained model. The proposed method and its variants are compared with two feature selection algorithms based on feature clustering. These algorithms were evaluated in two well known data mining problems: classification and clustering. The results obtained show that the proposed algorithm obtained good accuracy and computational efficiency results, additionally not requiring the definition of critical parameters by the user.