Selection Method Research Articles

An automatically recursive feature elimination method based on threshold decision in random forest classification

ABSTRACT The rich feature information contained in the diverse remote sensing data has also exhibited growing potential in the field of image classification. However, the processing of multi-feature data still grapples with the challenges posed by the “curse of dimensionality” and the high computational costs. This paper proposes a remote sensing feature parameter selection method. This method is based on the Gini index of a Random Forest (RF) classifier and employs a 10% threshold decision to identify the optimal combination of remote sensing feature parameters. First, spectral features, texture features, temperature features, elevation features, and principal component features are selected to create a stack of remote sensing images. Second, multiple sets of decision trees are established to cross-validate the contributions of various feature parameters. The feature rankings are determined based on the normalized mean of feature importance. Subsequently, a threshold isset to filter out remote sensing feature parameters that meet the criteria. Finally, an iterative parameter optimization process is carried out to obtain the optimal combination of remote sensing feature parameters. The Landsat 8 image covering Hangzhou Bay in eastern China and the Sentinel-2 remote sensing image of Yancheng Nature Reserve in Jiangsu, China were selected for experiments. The results show that the feature parameters of the remote sensing image screened by the method in this paper are representative, and the proposed method demonstrates strong adaptability and generalization performance in complex environments. The study has important guiding significance for regional spatial planning and sustainable development.

Forward variable selection enables fast and accurate dynamic system identification with Karhunen-Loève decomposed Gaussian processes

A promising approach for scalable Gaussian processes (GPs) is the Karhunen-Loève (KL) decomposition, in which the GP kernel is represented by a set of basis functions which are the eigenfunctions of the kernel operator. Such decomposed kernels have the potential to be very fast, and do not depend on the selection of a reduced set of inducing points. However KL decompositions lead to high dimensionality, and variable selection thus becomes paramount. This paper reports a new method of forward variable selection, enabled by the ordered nature of the basis functions in the KL expansion of the Bayesian Smoothing Spline ANOVA kernel (BSS-ANOVA), coupled with fast Gibbs sampling in a fully Bayesian approach. It quickly and effectively limits the number of terms, yielding a method with competitive accuracies, training and inference times for tabular datasets of low feature set dimensionality. Theoretical computational complexities are O ( N P 2 ) in training and O ( P ) per point in inference, where N is the number of instances and P the number of expansion terms. The inference speed and accuracy makes the method especially useful for dynamic systems identification, by modeling the dynamics in the tangent space as a static problem, then integrating the learned dynamics using a high-order scheme. The methods are demonstrated on two dynamic datasets: a ‘Susceptible, Infected, Recovered’ (SIR) toy problem, along with the experimental ‘Cascaded Tanks’ benchmark dataset. Comparisons on the static prediction of time derivatives are made with a random forest (RF), a residual neural network (ResNet), and the Orthogonal Additive Kernel (OAK) inducing points scalable GP, while for the timeseries prediction comparisons are made with LSTM and GRU recurrent neural networks (RNNs) along with the SINDy package.

Recent Advances inAgrobacterium-Mediated Maize Genetic Transformation Using Immature Embryos and Seedling Leaf Whorl Explants

The introduction of maize genetic transformation in the 1990s brought forth a powerful tool for crop improvement and a deeper understanding of plant genetics. Despite decades of genetics research, however, and the promise of CRISPR-mediated gene editing, maize transformation currently faces several challenges, such as genotype dependence and limitations in explant availability. Indeed, although the most commonly used method, immature embryo transformation, has been improved through optimization of tissue culture media composition and selection methods, the approach is only applicable to a limited number of public genotypes, including B104 and Hi II. Recently, genotype-flexible methods have been developed using coexpression cassettes of morphogenic transcription factors (MTFs)Baby boom(Bbm) andWushel2(Wus2), which have enabled the successful transformation of many previously recalcitrant maize lines. This MTF-based transformation method has also allowed for the use of alternate explants, such as seedling leaf whorl, whose production is cost-effective and requires only minimum controlled growth space. In this review, we summarize recent advances inAgrobacterium-mediated maize transformation methods that use immature embryos or seedling leaf whorls as starting material.

A Combined OCBA–AIC Method for Stochastic Variable Selection in Data Envelopment Analysis

This study introduces a novel approach to enhance variable selection in Data Envelopment Analysis (DEA), especially in stochastic environments where efficiency estimation is inherently complex. To address these challenges, we propose a game cross-DEA model to refine efficiency estimation. Additionally, we integrate the Akaike Information Criterion (AIC) with the Optimal Computing Budget Allocation (OCBA) technique, creating a hybrid method named OCBA–AIC. This innovative method efficiently allocates computational resources for stochastic variable selection. Our numerical analysis indicates that OCBA–AIC surpasses existing methods, achieving a lower AIC value. We also present two real-world case studies that demonstrate the effectiveness of our approach in ranking suppliers and tourism companies under uncertainty by selecting the most suitable partners. This research enriches the understanding of efficiency measurement in DEA and makes a substantial contribution to the field of performance management and decision-making in stochastic contexts.

Prevalence, satisfaction and preference of tooth shades and their correlation with age, gender and skin shade: A cross sectional study

Shade selection is one of the most important factors that determines the satisfaction of the patient with the dental prosthesis and thus the overall success of the treatment. Conventional methods of shade selection require the presence of existing teeth or past dental/photographic records. In the absence of these, shade selection becomes challenging and must be done subjectively and arbitrarily based on the patient’s age and sex. This study aims to assess the prevalence, satisfaction, and preference of tooth shades, investigating their correlation with age, sex, and skin shade in a diverse sample of 120 participants. Methods A total of 120 participants, evenly distributed among 60 males and 60 females, were surveyed. Participants were categorized into age groups (n=15) and asked about their satisfaction with existing tooth shades and their preferences. Results The most common existing tooth shade was Value 2 (52.5%). Statistical analysis using Chi-square tests revealed that Value 2 was the most desired shade among males and females alike, particularly in the age group of 41 to 50 years (p < 0.001). Shade 2L1.5 was the most prevalent (n=26), with Value 2 (59.16%) being the most sought-after shade, followed by Shade 2L1.5 (n=21), Shade 2M1 (n=18), Shade 1M1 (n=15), and Shade 2R1.5 (n=14). Existing tooth shades of Value 0 were the least common. The highest satisfaction with existing shades was observed in males aged 41 to 50 (n=15), followed by males aged 51 to 60 (n=14) and females aged 31 to 40 (n=12) and 41 to 50 (n=12). Significant differences in satisfaction were noted among males (p < 0.001). Conclusions This study reveals a consistent and strong preference among both males and females for tooth shades classified as Value 2, following the VITA Tooth-Guide 3D Master shade guide. This universal preference suggests the relevance and acceptance of this shade category across diverse patient groups.

Exploring Project Management Office Models for Public Construction Projects in Hail, Saudi Arabia

The Saudi Vision 2030 initiative is critical to boosting national development, particularly in the construction industry. The construction industry is getting a lot of attention in a variety of ways, the most visible of which is the promotion of best practices, including project management office (PMO) applications. Hail province, like others, is experiencing project delivery issues. Thus, the purpose of this research is to explore the most suitable PMO model for Hail public construction projects. A literature review included an examination of the driving factors for establishing PMOs in construction firms, as well as PMO models and typologies that describe their functions. Therefore, those driving factors and PMO models were examined in a focus group to explore their suitability in the context of Hail. The research findings revealed that the Enterprise PMO is the main model for Hail public construction projects, followed, respectively, by the Program Management Office model and the Project Management Center of Excellence (PMCOE). This method of selection for a PMO model encourages construction companies and related businesses to change in order to improve the sector while also addressing construction project delivery challenges in Hail province.

Forest Change Monitoring Based on Block Instance Sampling and Homomorphic Hypothesis Margin Evaluation

Forests play a crucial role in maintaining the integrity of natural ecosystems. Accurate mapping of windfall damages following storms is essential for effective post-disaster management. While remote sensing image classification offers substantial advantages over ground surveys for monitoring changes in forests, it encounters several challenges. Firstly, training samples in classification algorithms are typically selected through pixel-based random sampling or manual regional sampling. This approach struggles with accurately modeling complex patterns in high-resolution images and often results in redundant samples. Secondly, the limited availability of labeled samples compromises the classification accuracy when they are divided into training and test sets. To address these issues, two innovative approaches are proposed in this paper. The first is a new sample selection method which combines block-based sampling with spatial features extracted by single or multiple windows. Second, a new evaluation criterion is proposed by using the homomorphic hypothesis margin map with out-of-bag (OOB) accuracy. The former can not only assess the confidence level of each pixel category but also make regional boundaries clearer, and the latter can replace the test set so that all samples can be used for change detection. The experimental results show that the OOB accuracy obtained by spatial features with whole block sampling was 7.2% higher than that obtained by spectral features with pixel-based sampling and 2–3% higher than that for block center sampling, of which the highest value reached 98.8%. Additionally, the feasibility of identifying storm-damaged forests using only post-storm images has been demonstrated.

Time-Series Feature Selection for Solar Flare Forecasting

Solar flares are significant occurrences in solar physics, impacting space weather and terrestrial technologies. Accurate classification of solar flares is essential for predicting space weather and minimizing potential disruptions to communication, navigation, and power systems. This study addresses the challenge of selecting the most relevant features from multivariate time-series data, specifically focusing on solar flares. We employ methods such as Mutual Information (MI), Minimum Redundancy Maximum Relevance (mRMR), and Euclidean Distance to identify key features for classification. Recognizing the performance variability of different feature selection techniques, we introduce an ensemble approach to compute feature weights. By combining outputs from multiple methods, our ensemble method provides a more comprehensive understanding of the importance of features. Our results show that the ensemble approach significantly improves classification performance, achieving values 0.15 higher in True Skill Statistic (TSS) values compared to individual feature selection methods. Additionally, our method offers valuable insights into the underlying physical processes of solar flares, leading to more effective space weather forecasting and enhanced mitigation strategies for communication, navigation, and power system disruptions.

Incorporating Digital Footprints into Credit-Scoring Models through Model Averaging

Digital footprints provide crucial insights into individuals’ behaviors and preferences. Their role in credit scoring is becoming increasingly significant. Therefore, it is crucial to combine digital footprint data with traditional data for personal credit scoring. This paper proposes a novel credit-scoring model. First, lasso-logistic regression is used to select key variables that significantly impact the prediction results. Then, digital footprint variables are categorized based on business understanding, and candidate models are constructed from various combinations of these groups. Finally, the optimal weight is selected by minimizing the Kullback–Leibler loss. Subsequently, the final prediction model is constructed. Empirical analysis validates the advantages and feasibility of the proposed method in variable selection, coefficient estimation, and predictive accuracy. Furthermore, the model-averaging method provides the weights for each candidate model, providing managerial implications to identify beneficial variable combinations for credit scoring.

A Machine Learning-Based Wrapper Method for Feature Selection

This paper presents a two-stage feature selection scheme using machine learning techniques. In the first stage a wrapper method is adopted to select various combinations of subsets of features from the original dataset. The performance of the model is evaluated by three classifiers: K-Nearest Neighbor (KNN), Support Vector Machines (SVM), and Random Forest (RF). In the second and final stage, a sequential backward feature selection Method is applied. The proposed method is demonstrated on eighteen datasets and the average classification accuracy of eighteen datasets achieved is 89.81%, 87.55%, and 89.82% using the KNN, SVM, and RF classifiers, respectively with a maximum reduced size of the subset being ten only. Comparing the proposed method to eight other feature selection methods, the former achieves better classification accuracy in terms of selecting the most useful but a smaller number of features.

Magnetic resonance imaging-based machine learning classification of schizophrenia spectrum disorders: a meta-analysis.

Recent advances in multivariate pattern recognition have fostered the search for reliable neuroimaging-based biomarkers in psychiatric conditions, including schizophrenia. These approaches consider the complex pattern of alterations in brain function and structure, overcoming the limitations of traditional univariate methods. To assess the reliability of neuroimaging-based biomarkers and the contribution of study characteristics in distinguishing individuals with schizophrenia spectrum disorder (SSD) from healthy controls (HCs), we conducted a systematic review of the studies that used multivariate pattern recognition for this objective. We systematically searched PubMed, Scopus, and Web of Science for studies on SSD classification using multivariate pattern analysis on magnetic resonance imaging data. We employed a bivariate random-effects meta-analytic model to explore the classification of sensitivity (SE) and specificity (SP) across studies while also evaluating the moderator effects of clinical and non-clinical variables. A total of 119 studies (with 12,723 patients with SSD and 13,196 HCs) were identified. The meta-analysis estimated a SE of 79.1% (95% confidence interval [CI], 77.1%-81.0%) and a SP of 80.0% (95% CI, 77.8%-82.0%). In particular, the Positive and Negative Syndrome Scale and the Global Assessment of Functioning scores, age, age of onset, duration of untreated psychosis, deep learning, algorithm type, features selection, and validation methods had significant effects on classification performance. Multivariate pattern analysis reliably identifies neuroimaging-based biomarkers of SSD, achieving ∼80% SE and SP. Despite clinical heterogeneity, discernible brain modifications effectively differentiate SSD from HCs. Classification performance depends on patient-related and methodological factors crucial for the development, validation, and application of prospective models in clinical settings.

Enhancing deep learning-based slope stability classification using a novel metaheuristic optimization algorithm for feature selection

The evaluation of slope stability is of crucial importance in geotechnical engineering and has significant implications for infrastructure safety, natural hazard mitigation, and environmental protection. This study aimed to identify the most influential factors affecting slope stability and evaluate the performance of various machine learning models for classifying slope stability. Through correlation analysis and feature importance evaluation using a random forest regressor, cohesion, unit weight, slope height, and friction angle were identified as the most critical parameters influencing slope stability. This research assessed the effectiveness of machine learning techniques combined with modern feature selection algorithms and conventional feature analysis methods. The performance of deep learning models, including recurrent neural networks (RNNs), long short-term memory (LSTM) networks, and generative adversarial networks (GANs), in slope stability classification was evaluated. The GAN model demonstrated superior performance, achieving the highest overall accuracy of 0.913 and the highest area under the ROC curve (AUC) of 0.9285. Integration of the binary bGGO technique for feature selection with the GAN model led to significant improvements in classification performance, with the bGGO-GAN model showing enhanced sensitivity, positive predictive value, negative predictive value, and F1 score compared to the classical GAN model. The bGGO-GAN model achieved 95% accuracy on a substantial dataset of 627 samples, demonstrating competitive performance against other models in the literature while offering strong generalizability. This study highlights the potential of advanced machine learning techniques and feature selection methods for improving slope stability classification and provides valuable insights for geotechnical engineering applications.

Shielding networks: enhancing intrusion detection with hybrid feature selection and stack ensemble learning

The frequent usage of computer networks and the Internet has made computer networks vulnerable to numerous attacks, highlighting the critical need to enhance the precision of security mechanisms. One of the most essential measures to safeguard networking resources and infrastructures is an intrusion detection system (IDS). IDSs are widely used to detect, identify, and track malicious threats. Although various machine learning algorithms have been used successfully in IDSs, they are still suffering from low prediction performances. One reason behind the low accuracy of IDSs is that existing network traffic datasets have high computational complexities that are mainly caused by redundant, incomplete, and irrelevant features. Furthermore, standalone classifiers exhibit restricted classification performance and typically fail to produce satisfactory outcomes when dealing with imbalanced, multi-category traffic data. To address these issues, we propose an efficient intrusion detection model, which is based on hybrid feature selection and stack ensemble learning. Our hybrid feature selection method, called MI-Boruta, combines mutual information (MI) as a filter method and the Boruta algorithm as a wrapper method to determine optimal features from our datasets. Then, we apply stacked ensemble learning by using random forest (RF), Catboost, and XGBoost algorithms as base learners with multilayer perceptron (MLP) as meta-learner. We test our intrusion detection model on two widely recognized benchmark datasets, namely UNSW-NB15 and CICIDS2017. We show that our proposed IDS outperforms existing IDSs in almost all performance criteria, including accuracy, recall, precision, F1-Score, false positive rate, true positive rate, and error rate.

An Interpretable Deep-Stacked TSK Fuzzy Classifier for High-Dimensional Problems

This study proposes a novel deep-stacked fully interpretable and more generalization Takagi-Sugeno-Kang fuzzy system (D-FIMG-TSK) to obtain short fuzzy rules with high interpretability in each layer and simultaneously earn more generalization capability for high-dimensional data classification tasks. With the help of both discriminative and residual information, D-FIMG-TSK combines several fully interpretable TSK classifier FIMG-TSK in a stacked manner to conveniently move apart the manifolds existing in the original data space to achieve better linear separability. Besides, a feature selection method is designed to make feature selection from original feature sets on all layers to guarantee the classification performances. On the training process of D-FIMG-TSK, the importance of all original features and the antecedent and consequent parts of all fuzzy rules in each FIMG-TSK can be determined at the same time. The effectiveness of D-FIMG-TSK is manifested by the experimental results on eight binary high-dimensional classification datasets.

Development of Synchronization Selection Method in IoT with Secure Channel Bidirectional Communication

Background: The rapid development of wireless communications and mobile computation has given rise to the novel Internet of Things (IoT) systems, which is causing considerable research attention and industrial development. However, the lack of synchronization between the timers of IoT devices compromises the network's security. Aim: The purpose of this patent application is to present a technique for synchronizing the timepieces of IoT gadgets and establishing a secure channel for the transmission of data from source to destination. Objective: This study proposes a Synchronization Selection Method (SSM) for IoT systems to enhance network security and reduce packet loss. Methods: The method utilizes time-lay synchronization and RSA algorithm-based secure channel establishment. Time lay is a technique that was developed for IoT devices to achieve efficient clock synchronization of sensor nodes. Before synchronizing the sensor nodes' timings, the cluster leaders initiate the process. Utilizing a finite number of nodes, the proposed method was executed in MATLAB. Results: Time-lay synchronization involves all network nodes synchronizing their clocks with a third-party clock. In the context of time-lay synchronization, the term “third-party clock” refers to a single specific point that contains the time signal that all nodes in the network use as a reference. This third-party clock is outside of the network nodes and acts as the standard for the precise and synchronized time within the network. Therefore, it can be deduced that each of the techniques possesses its advantages and disadvantages. Each of the synchronization techniques has the potential to significantly benefit the IoT by offering smart clock synchronization that is more secure. Experimental results demonstrate that the proposed method improves throughput and reduces packet loss compared to existing techniques. Conclusion: The potential of this patent is highly significant for solving the synchronization problem of IoT devices and enhancing network security with decreased network packet loss. Other: The SSM would be assessed using the parameters of packet loss and throughput.

Reducing treatment time for robotic radiosurgery by considering path efficiency during node selection.

Robotic radiosurgery treatments allow for precise non-coplanar beam delivery by utilizing a robot equipped with a linac that traverses through a set of predetermined nodes. High quality treatment plans can be produced but treatment times can grow large, with one substantial component being the robot traversal time. The aim of this study is to reduce the treatment time for robotic radiosurgery treatments by introducing algorithms for reducing the robot traversal time. The algorithms are integrated into a commercial treatment planning system. First, an optimization framework for robotic radiosurgery planning is detailed, including a heuristic optimization method for node selection. Second, two methods aimed at reducing the traversal time are introduced. One utilizes a centrality measure focusing on the structure of the node network, while the other is based on the direct computation of traversal times during optimization. A comparison between plans with and without the time-reducing algorithms is made for three brain cases and one liver case with basis in treatment time, plan quality, monitor units, and network structure of the selectednodes. Large decreases in traversal times are obtained by the traversal time reducing algorithms, with reductions of up to 49 % in the brain cases and 31 % in the liver case. The resulting reductions in treatment times are up to 30 % and 13 %, respectively. Small differences in plan quality are observed, with similar dose-volume histograms, dose distributions, and conformity/gradientindices. The total treatment time of the robotic radiosurgery treatments can be reduced by selecting nodes with more efficient robot traversal paths, while maintaining planquality.

Highly efficient recovery of lithium from spent ternary NCM cathode materials via a selective chlorination method

The recovery of lithium from spent lithium-ion batteries (LIBs) is an important approach to solve the global shortage of lithium resource. Considering the difference in stability of various metal element chlorides in Ni-Co-Mn ternary (NCM) cathode materials, lithium element was selectively and high-efficiently extracted from spent ternary NCM cathode material by ammonium chloride roasting in oxygen atmosphere followed by water leaching. The effects of roasting temperature, roasting time, and NH4Cl dosage on lithium leaching efficiency and selectivity were systematically investigated. In addition, the conversion mechanism during chloride roasting was discussed in detail in conjunction with thermodynamic analysis as well as structural and compositional characterizations. After roasting with NH4Cl in oxygen atmosphere at 600°C, the spent NCM cathode material was thoroughly decomposed and converted into mixture of LiCl, NiO, and MnCo2O4. Due to the high aqueous solubility of LiCl, more than 95 % of the lithium in the mixture was leached in ∼30 minutes at room temperature, and Li2CO3 with a purity of 99.40 % can be obtained in the subsequent recovery process. This simple and efficient method can be generally useful for the recovery of a wide range of spent cathode materials and provide inspiration for the lithium recovery process of spent LIBs.

Effect of Differential Item Functioning on Computer Adaptive Testing Under Different Conditions

This study provides an overview of the effect of differential item functioning (DIF) on measurement precision, test information function (TIF), and test effectiveness in computer adaptive tests (CATs). Simulated data for the study was produced and analyzed with the Rstudio. During the data generation process, item pool size, DIF type, DIF percentage, item selection method for CAT, and the test termination rules were considered changed conditions. Sample size and ability parameter distribution, Item Response Theory (IRT) model, DIF size, ability estimation method, test starting rule, and item usage frequency method regarding CAT conditions were considered fixed conditions. To examine the effect of DIF, measurement precision, TIF and test effectiveness were calculated. Results show DIF has negative effects on measurement precision, TIF, and test effectiveness. In particular, statistically significant effects of the percentage DIF items and DIF type are observed on measurement precision.

Exploring the Impacts of Fescue Toxicosis on the Pulmonary Arterial Pressure of Angus Cows.

Vasoconstriction of peripheral blood vessels is one of the hallmark symptoms of fescue toxicosis in cattle. Thus, it was hypothesized that exposure to ergot alkaloids would increase the pulmonary arterial pressure (PAP). The objectives of this study were to examine the relationship between PAP and different physiological parameters of cows grazing either endophyte-infected (EI) or novel-endophyte (EN) fescue, then evaluate changes in PAP and other physiological measurements in cows exposed to EI pastures and deemed as susceptible or tolerant based on animal performance. Pregnant Angus cows at two different locations grazed either EI or EN fescue pastures for 14 consecutive weeks starting in early April of 2022. Forage measurements were collected to assess ergot alkaloid exposure throughout the study. In addition to measuring PAP, weekly measurements and blood samples were collected to evaluate physiological responses to ergot alkaloid consumption. The Fescue Toxicosis Selection Method (FTSM) was used for a post hoc analysis to identify cattle as either tolerant (EI-TOL) or susceptible (EI-SUS) when challenged with ergot alkaloid exposure. Data were analyzed using a MIXED procedure of SAS with repeated measures. Cows grazing on EN pastures had greater mean PAP values than EI cows, (P < 0.01), whereas a location effect was identified when comparing both EI-TOL and EI-SUS groups (P < 0.01). Cows exposed to EN pastures had greater ADG (P = 0.04) and progesterone (P4) concentrations (P < 0.01), and lower hair shedding scores (HSS; P < 0.01) than EI cows. The EI-TOL cows tended to have greater final BW, ADG, and had lower HSS (P < 0.01) than EI-SUS cows. While cattle consuming EI tall fescue exhibited classical physiological changes, the decrease in PAP of cattle consuming EI fescue was unexpected and contradicts the initial hypothesis. Furthermore, the FTSM provides a means to identify animals with superior performance in spite of the chronic exposure to ergot alkaloids. Continued investigations examining the interaction between ergot alkaloid exposure on cardiovascular parameters will lead to a fuller understanding of the disease, and are pivotal for developing innovative strategies that enhance best management practices to help guarantee the sustainability of the U.S. beef industry.

A novel optimal resonance band selection method for wheelset-bearing fault diagnosis based on tunable-Q wavelet transform

The detection of faults in the wheelset-bearing system is crucial for guaranteeing the safety of train operations. The core is to extract the optimal resonance band (ORB) and repetitive transient impact signals from the collected axle box acceleration signals. Accordingly, a novel automatic fault detection method called the bidirectional iterative merging multi-Q tunable-Q wavelet transform (BIMMQTQWT) is proposed to address the issue that existing methods are vulnerable to background noise and irrelevant components. First, a series of band-pass filters with almost constant bandwidth are constructed by the improved multi-Q tunable-Q wavelet transform (IMQTQWT) derived from the fault characteristics. Second, the fault information contained in each sub-band coefficient is preliminarily estimated using the correlative envelope comprehensive indicator (CECI). Third, the ORBs are automatically selected using the maximum CECI based on a strategy called bidirectionally merging adjacent frequency bands (BIMFBs). Finally, Envelope demodulation based on the ORB is executed followed by identifying bearing faults. The effectiveness in detecting multiple wheelset-bearing faults of the proposed method is validated through simulation and bench experiment signals. And the superior performance of the proposed method is exhibited compared with the existing average infogram and resonance sparse decomposition.