Existing Feature Selection Methods Research Articles

ScFSNN: a feature selection method based on neural network for single-cell RNA-seq data

While single-cell RNA sequencing (scRNA-seq) allows researchers to analyze gene expression in individual cells, its unique characteristics like over-dispersion, zero-inflation, high gene-gene correlation, and large data volume with many features pose challenges for most existing feature selection methods. In this paper, we present a feature selection method based on neural network (scFSNN) to solve classification problem for the scRNA-seq data. scFSNN is an embedded method that can automatically select features (genes) during model training, control the false discovery rate of selected features and adaptively determine the number of features to be eliminated. Extensive simulation and real data studies demonstrate its excellent feature selection ability and predictive performance.

Feature selection for hybrid information systems based on fuzzy β covering and fuzzy evidence theory

Feature selection can remove data noise and redundancy and reduce computational complexity, which is vital for machine learning. Because the difference between nominal attribute values is difficult to measure, feature selection for hybrid information systems faces challenges. In addition, many existing feature selection methods are susceptible to noise, such as Fisher, LASSO, random forest, mutual information, rough-set-based methods, etc. This paper proposes some techniques that consider the above problems from the perspective of fuzzy evidence theory. Firstly, a new distance incorporating decision attributes is defined, and then a relation between fuzzy evidence theory and fuzzy β covering with an anti-noise mechanism is established. Based on fuzzy belief and fuzzy plausibility, two robust feature selection algorithms for hybrid data are proposed in this framework. Experiments on 10 datasets of various types have shown that the proposed algorithms achieved the highest classification accuracy 11 times out of 20 experiments, significantly surpassing the performance of the other 6 state-of-the-art algorithms, achieved dimension reduction of 84.13% on seven UCI datasets and 99.90% on three large-scale gene datasets, and have a noise tolerance that is at least 6% higher than the other 6 state-of-the-art algorithms. Therefore, it can be concluded that the proposed algorithms have excellent anti-noise ability while maintaining good feature selection ability.

Defining a radiomics feature selection method for predicting response to transarterial chemoembolization in hepatocellular carcinoma patients

AimTo assess the utility of different radiomics feature selection methods in predicting transarterial chemoembolization (TACE) response in hepatocellular carcinoma (HCC) patients. Materials and methodsThis study employed a dataset of 136 paired MR T1-weighted contrast-enhanced abdominal images with liver tumor masks before and after TACE. TACE response for each image pair was classified by European Association for the Study of the Liver (EASL) and modified Response Evaluation Criteria in Solid Tumors (mRECIST) guidelines. 100D feature vectors were generated for the paired tumor areas. Eighteen existing feature selection methods were employed to select the top-k features to train and test a non-linear support vector machine (SVM) with a Gaussian kernel. Five-cross validation was performed to identify the highest performing feature selection methods. ResultsFor all benchmarks, a L0-based method selecting the top-5 or top-10 features achieved the highest performance. For images classified with EASL criteria that were analyzed with the L0-based method, the accuracy (ACC), area under curve (AUC), and balanced F score (F1-score) were 0.75 ​± ​0.06, 0.75 ​± ​0.09, and 0.80 ​± ​0.05, respectively. For images classified with mRECIST criteria that were analyzed with the L0-based method, the ACC, AUC, and F1-score were 0.75 ​± ​0.07, 0.71 ​± ​0.16, and 0.82 ​± ​0.04, respectively. ConclusionA L0-based method that selected the top-5/10 most important features predicted TACE response in HCC patients with the highest accuracy under both EASL and mRECIST criteria. This proof-of-concept investigation represents a step forward in the development of a reliable clinical decision-making tool for management of intermediate HCC patients undergoing TACE.

Feature importance feedback with Deep Q process in ensemble-based metaheuristic feature selection algorithms

Feature selection is an indispensable aspect of modern machine learning, especially for high-dimensional datasets where overfitting and computational inefficiencies are common concerns. Traditional methods often employ either filter, wrapper, or embedded approaches, which have limitations in terms of robustness, computational load, or capability to capture complex interactions among features. Despite the utility of metaheuristic algorithms like Particle Swarm Optimization (PSO), Firefly Algorithm (FA), and Whale Optimization (WOA) in feature selection, there still exists a gap in efficiently incorporating feature importance feedback into these processes. This paper presents a novel approach that integrates the strengths of PSO, FA, and WOA algorithms into an ensemble model and further enhances its performance by incorporating a Deep Q-Learning framework for relevance feedbacks. The Deep Q-Learning module intelligently updates feature importance based on model performance, thereby fine-tuning the selection process iteratively. Our ensemble model demonstrates substantial gains in effectiveness over traditional and individual metaheuristic approaches. Specifically, the proposed model achieved a 9.5% higher precision, an 8.5% higher accuracy, an 8.3% higher recall, a 4.9% higher AUC, and a 5.9% higher specificity across multiple software bug prediction datasets and samples. By resolving some of the key issues in existing feature selection methods and achieving superior performance metrics, this work paves the way for more robust and efficient machine learning models in various applications, from healthcare to natural language processing scenarios. This research provides an innovative framework for feature selection that promises not only superior performance but also offers a flexible architecture that can be adapted for a variety of machine learning challenges.

Research on bearing ZC feature selection method based on DWCMI

Abstract In light of the issues related to the omission of crucial features and the incorrect selection of redundant features in existing feature selection methods for zero-crossing features, this paper presents a feature selection method based on dynamic weights condition mutual information (DWCMI). In this method, the main factor of the objective function for feature evaluation is conditional mutual information, while also incorporating a complementary evaluation criterion based on conditional mutual information to address the issue of misselecting redundant features. By introducing a dynamic weight coefficient, we can accurately measure the importance of selected features by assessing their dynamic change in mutual information, thereby avoiding any oversight of crucial features. In the process of designing the algorithm, computational efficiency is improved by buffering and reusing previously calculated mutual information. This approach avoids the issue of repeatedly calculating the mutual information. The necessity, effectiveness, and high efficiency of the DWCMI method have been verified through simulation and experimentation.

Feature Selection Based on Intrusive Outliers Rather Than All Instances.

Feature selection (FS) has recently attracted considerable attention in many fields. Highly-overlapping classes and skewed distributions of data within classes have been found in various classification tasks. Most existing FS methods are all instance-based, which ignores the significant differences in characteristics between the particular outliers and the main body of the class, causing confusion for classifiers. In this paper, we propose a novel supervised FS method, Intrusive Outliers-based Feature Selection (IOFS), to find out what kind of outliers lead to misclassification and exploit the characteristics of such outliers. In order to accurately identify the intrusive outliers (IOs), we provide a density-mean center algorithm to obtain the appropriate representative of a class. A special distance threshold is given to obtain the candidate for IOs. Combining with several metrics, mathematical formulations are provided to evaluate the overlapping degree of the intrusive class pairs. Features with high overlapping degrees are assigned to low rankings in IOFS method. An extension of IOFS based on a small number of extreme IOs, called E-IOFS, is also proposed. Three theoretical proofs are provided for the essential theoretical basis of IOFS. Experiments comparing against various state-of-the-art methods on eleven benchmark datasets show that IOFS is rational and effective, especially on the datasets with higher overlapping classes. And E-IOFS almost always outperforms IOFS.

Unsupervised Discriminative Feature Selection via Contrastive Graph Learning.

Due to many unmarked data, there has been tremendous interest in developing unsupervised feature selection methods, among which graph-guided feature selection is one of the most representative techniques. However, the existing feature selection methods have the following limitations: (1) All of them only remove redundant features shared by all classes and neglect the class-specific properties; thus, the selected features cannot well characterize the discriminative structure of the data. (2) The existing methods only consider the relationship between the data and the corresponding neighbor points by Euclidean distance while neglecting the differences with other samples. Thus, existing methods cannot encode discriminative information well. (3) They adaptively learn the graph in the original or embedding space. Thus, the learned graph cannot characterize the data's cluster structure. To solve these limitations, we present a novel unsupervised discriminative feature selection via contrastive graph learning, which integrates feature selection and graph learning into a uniform framework. Specifically, our model adaptively learns the affinity matrix, which helps characterize the data's intrinsic and cluster structures in the original space and the contrastive learning. We minimize l1,2 -norm regularization on the projection matrix to preserve class-specific features and remove redundant features shared by all classes. Thus, the selected features encode discriminative information well and characterize the discriminative structure of the data. Generous experiments indicate that our proposed model has state-of-the-art performance.

Remote Parkinson's disease severity prediction based on causal game feature selection

Telemonitoring of Parkinson's disease has important implications for early diagnosis and treatment of patients. Most of the existing feature selection methods for remote prediction of PD severity are based on correlation and rarely consider causality, thus compromising the robustness of the model. Therefore, a causal game-based feature selection (CGFS) model is proposed for remote PD symptom severity assessment. Firstly, to address the challenge of small data size, the similar patient transfer strategy is designed to find data from source domain patients with conditions similar to those of the target patient. Secondly, the undirected equivalent greedy search method is employed to construct the causal graph between features and PD severity scores, and the robustness of the model is improved by selecting causal features. Then, to enhance the prediction accuracy, this paper utilizes the cooperative game approach Shapley value to evaluate the contribution of neighborhood nodes of the target value, and selects the features with causality and high contribution to form the final feature subset. Finally, the subset is input into the random forest to further enhance robustness and performance of the model. Experiments on Parkinson’s telemonitoring dataset and the tapping dataset with different biomarkers show that the robustness of the feature subset selected by the CGFS model, and the prediction performance is better than advanced models compared. Therefore, the validity and universality of the CGFS method is demonstrated in remote PD severity prediction.

MCDM-EFS: A novel ensemble feature selection method for software defect prediction using multi-criteria decision making

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.

Label correlations-based multi-label feature selection with label enhancement

Feature selection, as an important pre-processing technique, can efficiently mitigate the issue of “the curse of dimensionality” by selecting discriminative features especially for multi-label learning, a discriminative feature subset can improve the classification accuracy. The existing feature selection methods for multi-label classification address the problem of label ambiguity by with logical labels. However, the significance of each label is often different in many practical applications. Using logical label to train the model may result in unsatisfactory performance due to not considering the importance of related labels with each sample. To address this issue, a novel multi-label feature selection algorithm is proposed with two-step: label enhancement and label correlations-based feature selection with label enhancement. In the step of label enhancement, a framework of label enhancement based on deep forest is utilized to transform the logical label to label distribution, which contains rich semantic information and then guides a more correct exploration of semantic correlations. In the step of feature selection, a novel multi-label feature selection algorithm is proposed based on label distribution data. Firstly, the samples are divided into multiple different clusters by using spectral clustering in the label space. Then, the label correlations can be reflected by multiple different clusters. Finally, the l2,1-norm is used to construct an objective function to achieve multi-label feature selection. Experimental results demonstrate that competitiveness of the proposed algorithm over six state-of-the-art multi-label feature selection algorithms on eighteen benchmark datasets in terms of six widely accepted evaluation metrics.

Feature selection for single cell RNA sequencing data based on a noise-robust fuzzy relation and fuzzy evidence theory

Feature selection plays an important role in removing noise and redundancy of data. Due to various uncertainties in single cell RNA sequencing (scRNA-seq) data, such as noise, mutation, sparsity, etc., most of the existing feature selection methods for scRNA-seq data performed poorly. This paper has given some techniques that consider the above problems from the perspective of fuzzy evidence theory. First of all, a noise-robust fuzzy relation incorporating the decision attribute is defined because the decision attribute contains the most important information of data. Then, fuzzy evidence theory is introduced because it is good at handling the uncertainty of continuous data. Furthermore, the connection between fuzzy evidence theory and fuzzy rough set models has been established, which solves the problem of difficult calculation of fuzzy belief and fuzzy plausibility. In this framework, two noise-robust feature selection algorithms without any parameters for scRNA-seq data are proposed. Experiments on 12 scRNA-seq datasets show that the proposed algorithms are superior to the other four state-of-the-art algorithms in classification accuracy and noise-robust performance in the case of selecting fewer genes. Therefore, the proposed algorithms are noise-robust and objective while maintaining high gene selection efficiency and accuracy.

NNFSRR: Nearest Neighbor Feature Selection and Redundancy Removal Method for Nearest Neighbor Search in Microarray Gene Expression Data

INTRODUCTION: Gene expression data analysis is a critical aspect of disease prediction and classification, playing a pivotal role in the field of bioinformatics and biomedical research. High-dimensional gene expression datasets hold a wealth of information, but their effective utilization is hindered by the presence of irrelevant dimensions and noise. The challenge lies in extracting meaningful features from these datasets to enhance the accuracy of disease prediction and classification while maintaining computational efficiency.
 Feature selection is a crucial step in addressing these challenges, as it aims to identify and retain only the most informative characteristics from large high-dimensional microarray datasets. In the context of microarray gene expression data, characterized by its substantial dimensionality, selecting relevant features is essential for efficient nearest neighbor search, a fundamental component of various analytical tasks in bioinformatics and data mining.
 Existing feature selection methods in high-dimensional data often face issues related to the trade-off between search accuracy and computational efficiency. This paper introduces a novel approach, the Nearest Neighbor Feature Selection with Symmetrical Uncertainty-based Redundancy Removal (NNFSRR) method, designed to enhance the classification of microarray gene expression data through feature selection. The NNFSRR method focuses on reducing the dimensionality of the dataset by identifying and removing redundant features, allowing subsequent searches to operate solely on relevant dimensions.
 OBJECTIVES: The primary goal is to evaluate the NNFSRR method's effectiveness in improving nearest neighbor search in microarray gene expression datasets by reducing dimensionality. This method utilizes Symmetrical Uncertainty-based correlation between dimensions for feature selection and aims to enhance accuracy and efficiency compared to existing methods.
 METHODS: The NNFSRR method uses Symmetrical Uncertainty to identify and remove redundant features from microarray gene expression datasets. Reduced datasets are used for nearest neighbor search, improving accuracy and efficiency. Experiments are conducted using real-world datasets, and comparisons with existing methods are made based on search time and accuracy.
 RESULTS: The NNFSRR method demonstrates improved nearest neighbor search performance, outperforming basic brute force methods and existing feature selection techniques. Selected feature sets exhibit strong class associations while minimizing feature correlations, enhancing classification precision.
 CONCLUSION: In conclusion, the NNFSRR method presents a promising approach to address the challenges posed by high-dimensional gene expression data. It effectively reduces dimensionality, improves search accuracy, and enhances the efficiency of nearest neighbor search. Our experimental results demonstrate that this method outperforms existing techniques in terms of search time and accuracy, making it a valuable tool for applications in bioinformatics, data mining, pattern recognition, and biological information retrieval. The NNFSRR method holds the potential to advance our understanding of complex biological processes and support more accurate disease prediction and classification.

EEG feature selection method based on maximum information coefficient and quantum particle swarm

To reduce the dimensionality of EEG features and improve classification accuracy, we propose an improved hybrid feature selection method for EEG feature selection. First, MIC is used to remove irrelevant features and redundant features to reduce the search space of the second stage. QPSO is then used to optimize the feature in the second stage to obtain the optimal feature subset. Considering that both dimensionality and classification accuracy affect the performance of feature subsets, we design a new fitness function. Moreover, we optimize the parameters of the classifier while optimizing the feature subset to improve the classification accuracy and reduce the running time of the algorithm. Finally, experiments were performed on EEG and UCI datasets and compared with five existing feature selection methods. The results show that the feature subsets obtained by the proposed method have low dimensionality, high classification accuracy, and low computational complexity, which validates the effectiveness of the proposed method.

Supervised Relevance-Redundancy assessments for feature selection in omics-based classification scenarios

Background and objective:Many classification tasks in translational bioinformatics and genomics are characterized by the high dimensionality of potential features and unbalanced sample distribution among classes. This can affect classifier robustness and increase the risk of overfitting, curse of dimensionality and generalization leaks; furthermore and most importantly, this can prevent obtaining adequate patient stratification required for precision medicine in facing complex diseases, like cancer. Setting up a feature selection strategy able to extract only proper predictive features by removing irrelevant, redundant, and noisy ones is crucial to achieving valuable results on the desired task. Methods:We propose a new feature selection approach, called ReRa, based on supervised Relevance-Redundancy assessments. ReRa consists of a customized step of relevance-based filtering, to identify a reduced subset of meaningful features, followed by a supervised similarity-based procedure to minimize redundancy. This latter step innovatively uses a combination of global and class-specific similarity assessments to remove redundant features while preserving those differentiated across classes, even when these classes are strongly unbalanced. Results:We compared ReRa with several existing feature selection methods to obtain feature spaces on which performing breast cancer patient subtyping using several classifiers: we considered two use cases based on gene or transcript isoform expression. In the vast majority of the assessed scenarios, when using ReRa-selected feature spaces, the performances were significantly increased compared to simple feature filtering, LASSO regularization, or even MRmr — another Relevance-Redundancy method. The two use cases represent an insightful example of translational application, taking advantage of ReRa capabilities to investigate and enhance a clinically-relevant patient stratification task, which could be easily applied also to other cancer types and diseases. Conclusions:ReRa approach has the potential to improve the performance of machine learning models used in an unbalanced classification scenario. Compared to another Relevance-Redundancy approach like MRmr, ReRa does not require tuning the number of preserved features, ensures efficiency and scalability over huge initial dimensionalities and allows re-evaluation of all previously selected features at each iteration of the redundancy assessment, to ultimately preserve only the most relevant and class-differentiated features.

Uncertainty measurement for single cell RNA-seq data based on class-consistent technology with application to semi-supervised gene selection

Because of the high cost of label collection, people are now faced with a large number of partially labeled gene expression data (plge-data). Single cell RNA-seq data (scrs-data) are a kind of important plge-data and reflect the abundance of gene transcript mRNA measured directly or indirectly in cells. For convenience, a decision information system (DIS) based on scrs-data is called a single cell gene decision space (scgd-space). Due to the high dimensionality of scrs-data, feature selection must be done before clustering and classification. The existing feature selection methods based on equivalence relation are ineffective for the scgd-space owing to the strictness of equality between information values. To solve the above problems, this paper studies the uncertainty measurement of the scgd-space based on class-consistent technology and considers its application to semi-supervised gene selection. Class-consistent technology replaces equality with approximate equality between two expression values at a gene. Based on the proposed technology, class-consistent and non-class-consistent relations on the cell set of the scgd-space are established first. Then, the scgd-space (O,A,d) is divided into labeled space (Ol,A,d) and unlabeled space (Ou,A,d). Next, four metrics of importance on each gene subset of (O,A,d) are defined. They are the weighted sum of (Ol,A,d) and (Ou,A,d) determined by the missing rate of labels and the established relations and can be used to measure the uncertainty of (O,A,d). In addition, as an application of four metrics to the scgd-space, a semi-supervised gene selection algorithm is designed. Finally, the experimental results and statistical tests on 16 large-scale scrs-data sets show that the defined metrics can effectively measure the uncertainty of the scgd-space. The designed algorithm with a high reduction rate outperforms some state-of-the-art feature selection algorithms in terms of eight performance evaluation indicators.

Multiobjective Differential Evolution for Feature Selection in Classification.

Feature selection aims to reduce the number of features and improve the classification accuracy, which is an essential step in many real-world problems. Multiple feature subsets with different features selected can achieve similar or even the same objective values (e.g., maximize the classification accuracy and minimize the number of selected features). This means the optimal feature subsets of a classification problem may not be unique. However, most existing feature selection methods do not take into consideration finding multiple optimal feature subsets. In this article, a multiobjective differential evolution approach is developed to search for multiple optimal feature subsets. The contributions are three-fold. First, to provide a good starting point, an initialization method considering feature relevance is proposed. Second, a clustering method is used to divide the whole population into multiple subpopulations. In each of these subpopulations, a subarchive utilizes a developed crowding distance to ensure diversity by considering both the search space and the objective space. Finally, the nondominated solutions from all the subarchives are retained in another archive to guide the evolutionary feature selection process, together with an improved hypervolume contribution indicator. The experiments on 14 datasets of varying difficulty show that the proposed approach can evolve a better Pareto front of feature subsets compared with seven other state-of-the-art methods as well as find different feature subsets with similar or the same classification performance.

CellBRF: a feature selection method for single-cell clustering using cell balance and random forest.

Single-cell RNA sequencing (scRNA-seq) offers a powerful tool to dissect the complexity of biological tissues through cell sub-population identification in combination with clustering approaches. Feature selection is a critical step for improving the accuracy and interpretability of single-cell clustering. Existing feature selection methods underutilize the discriminatory potential of genes across distinct cell types. We hypothesize that incorporating such information could further boost the performance of single cell clustering. We develop CellBRF, a feature selection method that considers genes' relevance to cell types for single-cell clustering. The key idea is to identify genes that are most important for discriminating cell types through random forests guided by predicted cell labels. Moreover, it proposes a class balancing strategy to mitigate the impact of unbalanced cell type distributions on feature importance evaluation. We benchmark CellBRF on 33 scRNA-seq datasets representing diverse biological scenarios and demonstrate that it substantially outperforms state-of-the-art feature selection methods in terms of clustering accuracy and cell neighborhood consistency. Furthermore, we demonstrate the outstanding performance of our selected features through three case studies on cell differentiation stage identification, non-malignant cell subtype identification, and rare cell identification. CellBRF provides a new and effective tool to boost single-cell clustering accuracy. All source codes of CellBRF are freely available at https://github.com/xuyp-csu/CellBRF.

Unsupervised feature selection based on variance–covariance subspace distance

Subspace distance is an invaluable tool exploited in a wide range of feature selection methods. The power of subspace distance is that it can identify a representative subspace, including a group of features that can efficiently approximate the space of original features. On the other hand, employing intrinsic statistical information of data can play a significant role in a feature selection process. Nevertheless, most of the existing feature selection methods founded on the subspace distance are limited in properly fulfilling this objective. To pursue this void, we propose a framework that takes a subspace distance into account which is called “Variance–Covariance subspace distance”. The approach gains advantages from the correlation of information included in the features of data, thus determines all the feature subsets whose corresponding Variance–Covariance matrix has the minimum norm property. Consequently, a novel, yet efficient unsupervised feature selection framework is introduced based on the Variance–Covariance distance to handle both the dimensionality reduction and subspace learning tasks. The proposed framework has the ability to exclude those features that have the least variance from the original feature set. Moreover, an efficient update algorithm is provided along with its associated convergence analysis to solve the optimization side of the proposed approach. An extensive number of experiments on nine benchmark datasets are also conducted to assess the performance of our method from which the results demonstrate its superiority over a variety of state-of-the-art unsupervised feature selection methods. The source code is available at https://github.com/SaeedKarami/VCSDFS.

Low-rank learning for feature selection in multi-label classification

Considering several research areas such as data mining, pattern recognition, and machine learning, it is difficult to handle high-dimensional data sets because of time consumption, large memory, and inefficient algorithms. Therefore, the feature selection method is widely used to address these problems. Several conventional feature selection methods have selected features by calculating the relationship between features and labels using some measures such as a mutual information. However, it is difficult to calculate mutual information because it requires to joint probability from a high-dimensional data set. Therefore, only few features can be considered, and a global search cannot be used. Moreover, calculating the correlation among labels using mutual information is limited because of insufficient label sets. In this study, we propose a feature selection method for multi-label classification based on low-rank learning. Unlike conventional methods, which use restricted search space, the proposed method considers a low-rank space from the entire feature space provided and the relationship between features and labels simultaneously. Therefore, we design a regression-based objective function using a nuclear norm and propose an efficient algorithm based on the gradient descent method to solve the optimization problem. The proposed method shows a better performance than the existing feature selection methods. This is based on the results of the multi-label classification experiments with five data sets and five multi-label classification performances.

Multivariate sample similarity measure for feature selection with a resemblance model

Feature selection improves the classification performance of machine learning models. It also identifies the important features and eliminates those with little significance. Furthermore, feature selection reduces the dimensionality of training and testing data points. This study proposes a feature selection method that uses a multivariate sample similarity measure. The method selects features with significant contributions using a machine-learning model. The multivariate sample similarity measure is evaluated using the University of California, Irvine heart disease dataset and compared with existing feature selection methods. The multivariate sample similarity measure is evaluated with metrics such as minimum subset selected, accuracy, F1-score, and area under the curve (AUC). The results show that the proposed method is able to diagnose chest pain, thallium scan, and major vessels scanned using X-rays with a high capability to distinguish between healthy and heart disease patients with a 99.6% accuracy.