Random Forest Based Feature Selection Research Articles

Lipids balance as a spectroscopy marker of diabetes. Analysis of FTIR spectra by 2D correlation and machine learning analyses

The number of people suffering from type 2 diabetes has rapidly increased. Taking into account, that elevated intracellular lipid concentrations, as well as their metabolism, are correlated with diminished insulin sensitivity, in this study we would like to show lipids spectroscopy markers of diabetes. For this purpose, serum collected from rats (animal model of diabetes) was analyzed using Fourier Transformed Infrared-Attenuated Total Reflection (FTIR-ATR) spectroscopy. Analyzed spectra showed that rats with diabetes presented higher concentration of phospholipids and cholesterol in comparison with non-diabetic rats. Moreover, the analysis of second (IInd) derivative spectra showed no structural changes in lipids. Machine learning methods showed higher accuracy for IInd derivative spectra (from 65 % to 89 %) than for absorbance FTIR spectra (53–65 %). Moreover, it was possible to identify significant wavelength intervals from IInd derivative spectra using random forest-based feature selection algorithm, which further increased the accuracy of the classification (up to 92 % for phospholipid region). Moreover decision tree based on the selected features showed, that peaks at 1016 cm−1 and 2936 cm−1 can be good candidates of lipids marker of diabetes.

Comparison of cerebrospinal fluid, plasma and neuroimaging biomarker utility in Alzheimer's disease.

Alzheimer's disease biomarkers are crucial to understanding disease pathophysiology, aiding accurate diagnosis and identifying target treatments. Although the number of biomarkers continues to grow, the relative utility and uniqueness of each is poorly understood as prior work has typically calculated serial pairwise relationships on only a handful of markers at a time. The present study assessed the cross-sectional relationships among 27 Alzheimer's disease biomarkers simultaneously and determined their ability to predict meaningful clinical outcomes using machine learning. Data were obtained from 527 community-dwelling volunteers enrolled in studies at the Charles F. and Joanne Knight Alzheimer Disease Research Center at Washington University in St Louis. We used hierarchical clustering to group 27 imaging, CSF and plasma measures of amyloid beta, tau [phosphorylated tau (p-tau), total tau t-tau)], neuronal injury and inflammation drawn from MRI, PET, mass-spectrometry assays and immunoassays. Neuropsychological and genetic measures were also included. Random forest-based feature selection identified the strongest predictors of amyloid PET positivity across the entire cohort. Models also predicted cognitive impairment across the entire cohort and in amyloid PET-positive individuals. Four clusters emerged reflecting: core Alzheimer's disease pathology (amyloid and tau), neurodegeneration, AT8 antibody-associated phosphorylated tau sites and neuronal dysfunction. In the entire cohort, CSF p-tau181/Aβ40lumi and Aβ42/Aβ40lumi and mass spectrometry measurements for CSF pT217/T217, pT111/T111, pT231/T231 were the strongest predictors of amyloid PET status. Given their ability to denote individuals on an Alzheimer's disease pathological trajectory, these same markers (CSF pT217/T217, pT111/T111, p-tau/Aβ40lumi and t-tau/Aβ40lumi) were largely the best predictors of worse cognition in the entire cohort. When restricting analyses to amyloid-positive individuals, the strongest predictors of impaired cognition were tau PET, CSF t-tau/Aβ40lumi, p-tau181/Aβ40lumi, CSF pT217/217 and pT205/T205. Non-specific CSF measures of neuronal dysfunction and inflammation were poor predictors of amyloid PET and cognitive status. The current work utilized machine learning to understand the interrelationship structure and utility of a large number of biomarkers. The results demonstrate that, although the number of biomarkers has rapidly expanded, many are interrelated and few strongly predict clinical outcomes. Examining the entire corpus of available biomarkers simultaneously provides a meaningful framework to understand Alzheimer's disease pathobiological change as well as insight into which biomarkers may be most useful in Alzheimer's disease clinical practice and trials.

A review of random forest-based feature selection methods for data science education and applications

A review of random forest-based feature selection methods for data science education and applications

Identifying potential circulating miRNA biomarkers for the diagnosis and prediction of ovarian cancer using machine-learning approach: application of Boruta.

In gynecologic oncology, ovarian cancer is a great clinical challenge. Because of the lack of typical symptoms and effective biomarkers for noninvasive screening, most patients develop advanced-stage ovarian cancer by the time of diagnosis. MicroRNAs (miRNAs) are a type of non-coding RNA molecule that has been linked to human cancers. Specifying diagnostic biomarkers to determine non-cancer and cancer samples is difficult. By using Boruta, a novel random forest-based feature selection in the machine-learning techniques, we aimed to identify biomarkers associated with ovarian cancer using cancerous and non-cancer samples from the Gene Expression Omnibus (GEO) database: GSE106817. In this study, we used two independent GEO data sets as external validation, including GSE113486 and GSE113740. We utilized five state-of-the-art machine-learning algorithms for classification: logistic regression, random forest, decision trees, artificial neural networks, and XGBoost. Four models discovered in GSE113486 had an AUC of 100%, three in GSE113740 with AUC of over 94%, and four in GSE113486 with AUC of over 94%. We identified 10 miRNAs to distinguish ovarian cancer cases from normal controls: hsa-miR-1290, hsa-miR-1233-5p, hsa-miR-1914-5p, hsa-miR-1469, hsa-miR-4675, hsa-miR-1228-5p, hsa-miR-3184-5p, hsa-miR-6784-5p, hsa-miR-6800-5p, and hsa-miR-5100. Our findings suggest that miRNAs could be used as possible biomarkers for ovarian cancer screening, for possible intervention.

Enhancing Feature Selection for Imbalanced Alzheimer’s Disease Brain MRI Images by Random Forest

Imbalanced learning problems often occur in application scenarios and are additionally an important research direction in the field of machine learning. Traditional classifiers are substantially less effective for datasets with an imbalanced distribution, especially for high-dimensional longitudinal data structures. In the medical field, the imbalance of data problem is more common, and correctly identifying samples of the minority class can obtain important information. Moreover, class imbalance in imbalanced AD (Alzheimer’s disease) data presents a significant challenge for machine learning algorithms that assume the data are evenly distributed within the classes. In this paper, we propose a random forest-based feature selection algorithm for imbalanced neuroimaging data classification. The algorithm employs random forest to evaluate the value of each feature and combines the correlation matrix to choose the optimal feature subset, which is applied to imbalanced MRI (magnetic resonance imaging) AD data to identify AD, MCI (mild cognitive impairment), and NC (normal individuals). In addition, we extract multiple features from AD images that can represent 2D and 3D brain information. The effectiveness of the proposed method is verified by the experimental evaluation using the public ADNI (Alzheimer’s neuroimaging initiative) dataset, and results demonstrate that the proposed method has a higher prediction accuracy and AUC (area under the receiver operating characteristic curve) value in NC-AD, MCI-AD, and NC-MCI group data, with the highest accuracy and AUC value for the NC-AD group data.

Improved prediction of glioma-related aphasia by diffusion MRI metrics, machine learning, and automated fiber bundle segmentation.

White matter impairments caused by gliomas can lead to functional disorders. In this study, we predicted aphasia in patients with gliomas infiltrating the language network using machine learning methods. We included 78 patients with left-hemispheric perisylvian gliomas. Aphasia was graded preoperatively using the Aachen aphasia test (AAT). Subsequently, we created bundle segmentations based on automatically generated tract orientation mappings using TractSeg. To prepare the input for the support vector machine (SVM), we first preselected aphasia-related fiber bundles based on the associations between relative tract volumes and AAT subtests. In addition, diffusion magnetic resonance imaging (dMRI)-based metrics [axial diffusivity (AD), apparent diffusion coefficient (ADC), fractional anisotropy (FA), and radial diffusivity (RD)] were extracted within the fiber bundles' masks with their mean, standard deviation, kurtosis, and skewness values. Our model consisted of random forest-based feature selection followed by an SVM. The best model performance achieved 81% accuracy (specificity = 85%, sensitivity = 73%, and AUC = 85%) using dMRI-based features, demographics, tumor WHO grade, tumor location, and relative tract volumes. The most effective features resulted from the arcuate fasciculus (AF), middle longitudinal fasciculus (MLF), and inferior fronto-occipital fasciculus (IFOF). The most effective dMRI-based metrics were FA, ADC, and AD. We achieved a prediction of aphasia using dMRI-based features and demonstrated that AF, IFOF, and MLF were the most important fiber bundles for predicting aphasia in this cohort.

Alzheimer’s Disease Detection from Fused PET and MRI Modalities Using an Ensemble Classifier

Alzheimer’s disease (AD) is an old-age disease that comes in different stages and directly affects the different regions of the brain. The research into the detection of AD and its stages has new advancements in terms of single-modality and multimodality approaches. However, sustainable techniques for the detection of AD and its stages still require a greater extent of research. In this study, a multimodal image-fusion method is initially proposed for the fusion of two different modalities, i.e., PET (Positron Emission Tomography) and MRI (Magnetic Resonance Imaging). Further, the features obtained from fused and non-fused biomarkers are passed to the ensemble classifier with a Random Forest-based feature selection strategy. Three classes of Alzheimer’s disease are used in this work, namely AD, MCI (Mild Cognitive Impairment) and CN (Cognitive Normal). In the resulting analysis, the Binary classifications, i.e., AD vs. CN and MCI vs. CN, attained an accuracy (Acc) of 99% in both cases. The class AD vs. MCI detection achieved an adequate accuracy (Acc) of 91%. Furthermore, the Multi Class classification, i.e., AD vs. MCI vs. CN, achieved 96% (Acc).

An enhanced ensemble deep random vector functional link network for driver fatigue recognition

This work investigated the use of an ensemble deep random vector functional link (edRVFL) network for electroencephalogram (EEG)-based driver fatigue recognition. Against the low feature learning capability of the edRVFL network from raw EEG signals, two strategies were exploited in this work. Specifically, the first one was to exploit the advantages of the feature extractor module in CNNs, i.e., use CNN features as the input of the edRVFL network. The second one was to improve the feature learning capability of the edRVFL network. An enhanced edRFVL network named FGloWD-edRVFL was proposed, in which four enhancements were implemented, including random forest-based Feature selection, Global output layer, Weighting and entropy-based Dynamic ensemble. The proposed FGloWD-edRVFL network was evaluated on the challenging cross-subject driver fatigue recognition tasks. The results indicated that the proposed model could boost the recognition performance, significantly outperforming all strong baselines. The step-wise analysis further demonstrated the effectiveness of the proposed enhancements in the edRVFL network.

Prediction of heating and cooling loads based on light gradient boosting machine algorithms

Machine learning models have been widely used to study the prediction of heating and cooling loads in residential buildings. However, most of these methods use the default hyperparameters, resulting in inaccurate prediction accuracy. In this work, based on hyperparametric optimization algorithms of random search (Random), grid search (Grid), covariance matrix adaptive evolution strategy (CMA-ES), and tree-structured parzen estimator (TPE), were combined with the light gradient boosting machine (LightGBM) model, to construct four hybrid models (Random-LightGBM, Grid-LightGBM, CMA-ES-LightGBM and TPE-LightGBM) for improved prediction accuracy of heating and cooling loads. The LightGBM model was trained using a dataset consisting of building features, cooling set points, and occupant behavior parameters. Feature selection was performed by a random forest-based feature selection method, which determines the input features of the load prediction model. The TPE- LightGBM model achieved the best prediction accuracy among all proposed models with a root mean square error (RMSE) of 0.2714, mean absolute error (MAE) of 0.1416, coefficient of determination (R2) of 0.9981, and mean absolute percentage error (MAPE) of 0.4699% for heating load prediction, and RMSE of 0.1901, MAE of 0.1394, R2 of 0.9924, and MAPE of 2.3509% for cooling load prediction. The proposed TPE-LightGBM model provides an efficient strategy for predicting heating and cooling loads, which can provide better energy efficiency measures at the early design stages of residential buildings.

A hybrid random forest-based feature selection model using mutual information and F-score for preterm birth classification

A hybrid random forest-based feature selection model using mutual information and F-score for preterm birth classification

Research on the Application of Random Forest-based Feature Selection Algorithm in Data Mining Experiments

Research on the Application of Random Forest-based Feature Selection Algorithm in Data Mining Experiments

Immuno-AML, a Novel Immunogenomic Classifier Predicting Chemotherapy Response in Pediatric Patients with AML

Introduction Acute myeloid leukemia (AML) represents 20% of pediatric leukemias. Despite therapeutic improvements, ∼30% of children die due to therapy-related toxicity and relapse. In solid tumors, the density and functional orientation of the tumor microenvironment (TME) influence prognosis and response to chemotherapy and immunotherapy. Defining immunologic features associated with unfavorable response to standard treatment in pediatric AML might improve stratification systems that can be used for experimental interventions and personalized approaches. In this study, we investigated the potential mechanisms and effect of immune response and evasion in bone marrow (BM) of AML pediatric patients through an integrative gene expression analysis. Methods Thirty-two BM aspirates were collected from pediatric patients with non-promyelocytic AML at diagnosis in 3 different hospitals in Qatar, Pakistan, and Italy. Patients were assigned to 2 prognosis-related groups: patients who maintained a complete remission (CR) after first line of chemotherapy (sustained CR) vs patients who progressed/relapsed and/or obtained CR after hematopoietic stem cells transplantation (HSCT) (non-sustained CR). BM mRNA samples were analyzed with RNA sequencing (RNA-Seq) and/or NanoString's PanCancer IO 360 Gene Expression research panel. In particular, mRNA of 25 samples (Cohort 1, discovery set) were analyzed by NanoString's PanCancer IO 360. Subsequently, a subset of samples from Cohort 1 (n=19), was analyzed by RNA-Seq for technical validation. Cohort 1 results were then used to build a predictive classifier that was tested on Cohort 2 samples (n=7, validation set), analyzed with RNA-Seq. Results Gene expression analysis of 25 samples (Cohort 1, sustained CR, n=6; non-sustained CR, n=19) using NanoStringassay revealed 67 significant differentially expressed genes (DEGs; p < 0.05, LIMMA) among sustained CR vs non-sustained CR groups. Patients with sustained CR displayed coherent overexpression of immune genes related to IFN activation (IFIT3, IFITM1, IFI27, STAT1, GBP1), cytotoxic function (PRF1, GNLY), antigen presentation and B cell functions (TAP1, CD19), and T cell infiltration (CD3D, CD8A), consistent with a favorable T helper 1 (Th1)-cytotoxic polarization of the TME. This pattern was not observed in non-sustained CR patients, which overexpressed genes consistent with macrophage-myeloid suppression (CD68, TREM1) and intrinsic oncogene signaling linked with T cell exclusion and immune suppression (WNT5A, WNT-β- catenin pathway, and EFGR) (Fig1 A,B). This 67-gene signature (67GenSig) was technically validated using an orthogonal platform (RNA-Seq) on a subset of 19 patients (sustained CR, n=5; non-sustained CR, n=14) for which sufficient RNA-Seq data was available (65 out of 67 transcripts were available for analysis, 2 transcripts were filtered out after QC). Unsupervised PCA (Fig1 C) based on 67GenSig clearly segregated samples according to outcome, confirming the previous NanoString results. Next, we sought to identify a predictive classifier to precisely discriminate sustained CR from non-sustained CR AML pediatric patients using a minimal number of genes. By applying a Random Forest-based feature selection on 67GenSig transcripts of RNA-Seq data from Cohort 1 samples, we developed a 13-gene classifier (Immuno-AML) that was tested in a new independent dataset (Fig1 D; Cohort 2, n=7: sustained CR, n=4; non-sustained CR, n=3). Overall, genes included in the Immuno-AML predictor substantiated a positive Th1/cytotoxic vs myeloid polarization. The Immuno-AML predictor was able to classify patients in the validation cohort (Cohort 2) in "sustained CR" and "non-sustained CR" prognostic groups with a very good performance achieving 85.71% accuracy, 100% specificity, and 75% sensitivity. Conclusion Our data suggest that response to conventional therapy in AML pediatric patients is conditional to the presence of a Th1 polarization of the TME while is prevented by myeloid-mediated immune suppression and intrinsic oncogenic signaling. The optimized predictive signature (Immuno-AML) might be tested and/or refined in larger cohort to develop a clinical-grade predictive biomarker. It might help to provide a rational guidance for risk stratification criteria and personalized treatment decision paths, implying the possibility of combining immunotherapy drugs. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

A clustering-based sampling method for miRNA-disease association prediction.

More and more studies have proved that microRNAs (miRNAs) play a critical role in gene expression regulation, and the irregular expression of miRNAs tends to be associated with a variety of complex human diseases. Because of the high cost and low efficiency of identifying disease-associated miRNAs through biological experiments, scholars have focused on predicting potential disease-associated miRNAs by computational methods. Considering that the existing methods are flawed in constructing negative sample set, we proposed a clustering-based sampling method for miRNA-disease association prediction (CSMDA). Firstly, we integrated multiple similarity information of miRNA and disease to represent miRNA-disease pairs. Secondly, we performed a clustering-based sampling method to avoid introducing potential positive samples when constructing negative sample set. Thirdly, we employed a random forest-based feature selection method to reduce noise and redundant information in the high-dimensional feature space. Finally, we implemented an ensemble learning framework for predicting miRNA-disease associations by soft voting. The Precision, Recall, F1-score, AUROC and AUPR of the CSMDA achieved 0.9676, 0.9545, 0.9610, 0.9928, and 0.9940, respectively, under five-fold cross-validation. Besides, case study on three cancers showed that the top 20 potentially associated miRNAs predicted by the CSMDA were confirmed by the dbDEMC database or literatures. The above results demonstrate that the CSMDA can predict potential disease-associated miRNAs more accurately.

Metabolic drift in the aging nervous system is reflected in human cerebrospinal fluid

Chronic diseases affecting the central nervous system (CNS) like Alzheimer’s or Parkinson’s disease typically develop with advanced chronological age. Yet, aging at the metabolic level has been explored only sporadically in humans using biofluids in close proximity to the CNS such as the cerebrospinal fluid (CSF). We have used an untargeted liquid chromatography high-resolution mass spectrometry (LC-HRMS) based metabolomics approach to measure the levels of metabolites in the CSF of non-neurological control subjects in the age of 20 up to 74. Using a random forest-based feature selection strategy, we extracted 69 features that were strongly related to age (page < 0.001, rage = 0.762, R2Boruta age = 0.764). Combining an in-house library of known substances with in silico chemical classification and functional semantic annotation we successfully assigned putative annotations to 59 out of the 69 CSF metabolites. We found alterations in metabolites related to the Cytochrome P450 system, perturbations in the tryptophan and kynurenine pathways, metabolites associated with cellular energy (NAD+, ADP), mitochondrial and ribosomal metabolisms, neurological dysfunction, and an increase of adverse microbial metabolites. Taken together our results point at a key role for metabolites found in CSF related to the Cytochrome P450 system as most often associated with metabolic aging.

Intelligent selection and optimization method of feature variables in fluid catalytic cracking gasoline refining process

To suppress the research octane number (RON) loss in the gasoline refining process, an intelligent selection and optimization method of feature variables is proposed. In the methodology, the random forest-based feature selection algorithm is first used to calculate the importance of each feature variable, so that the main variables are selected for the prediction modeling of the RON loss and sulfur content. Next, the long short-term memory network is introduced to establish the nonlinear mapping relationship between the main feature variables and the product yield. Finally, an objective function for minimizing RON loss under the constraints of RON loss reduction and product sulfur content is constructed, and the optimal values of feature variables are obtained by the continuous iterative solution with a hybrid grey wolf optimizer algorithm. Based on real-world data in industrial processes, experimental results verify the feasibility and effectiveness of the proposed method.

Machine learning, transcriptome, and genotyping chip analyses provide insights into SNP markers identifying flower color in Platycodon grandiflorus

Bellflower is an edible ornamental gardening plant in Asia. For predicting the flower color in bellflower plants, a transcriptome-wide approach based on machine learning, transcriptome, and genotyping chip analyses was used to identify SNP markers. Six machine learning methods were deployed to explore the classification potential of the selected SNPs as features in two datasets, namely training (60 RNA-Seq samples) and validation (480 Fluidigm chip samples). SNP selection was performed in sequential order. Firstly, 96 SNPs were selected from the transcriptome-wide SNPs using the principal compound analysis (PCA). Then, 9 among 96 SNPs were later identified using the Random forest based feature selection method from the Fluidigm chip dataset. Among six machines, the random forest (RF) model produced higher classification performance than the other models. The 9 SNP marker candidates selected for classifying the flower color classification were verified using the genomic DNA PCR with Sanger sequencing. Our results suggest that this methodology could be used for future selection of breeding traits even though the plant accessions are highly heterogeneous.

Machine learning-based analysis of adolescent gambling factors.

Problem gambling among adolescents has recently attracted attention because of easy access to gambling in online environments and its serious effects on adolescent lives. We proposed a machine learning-based analysis method for predicting the degree of problem gambling. Of the 17,520 respondents in the 2018 National Survey on Youth Gambling Problems dataset (collected by the Korea Center on Gambling Problems), 5,045 students who had gambled in the past 3 months were included in this study. The Gambling Problem Severity Scale was used to provide the binary label information. After the random forest-based feature selection method, we trained four models: random forest (RF), support vector machine (SVM), extra trees (ETs), and ridge regression. The online gambling behavior in the past 3 months, experience of winning money or goods, and gambling of personal relationship were three factors exhibiting the high feature importance. All four models demonstrated an area under the curve (AUC) of >0.7; ET showed the highest AUC (0.755), RF demonstrated the highest accuracy (71.8%), and SVM showed the highest F1 score (0.507) on a testing set. The results indicate that machine learning models can convey meaningful information to support predictions regarding the degree of problem gambling. Machine learning models trained using important features showed moderate accuracy in a large-scale Korean adolescent dataset. These findings suggest that the method will help screen adolescents at risk of problem gambling. We believe that expandable machine learning-based approaches will become more powerful as more datasets are collected.

Machine learning-based analysis of adolescent gambling factors

Background and aimsProblem gambling among adolescents has recently attracted attention because of easy access to gambling in online environments and its serious effects on adolescent lives. We proposed a machine learning-based analysis method for predicting the degree of problem gambling.MethodsOf the 17,520 respondents in the 2018 National Survey on Youth Gambling Problems dataset (collected by the Korea Center on Gambling Problems), 5,045 students who had gambled in the past 3 months were included in this study. The Gambling Problem Severity Scale was used to provide the binary label information. After the random forest-based feature selection method, we trained four models: random forest (RF), support vector machine (SVM), extra trees (ETs), and ridge regression.ResultsThe online gambling behavior in the past 3 months, experience of winning money or goods, and gambling of personal relationship were three factors exhibiting the high feature importance. All four models demonstrated an area under the curve (AUC) of >0.7; ET showed the highest AUC (0.755), RF demonstrated the highest accuracy (71.8%), and SVM showed the highest F1 score (0.507) on a testing set.DiscussionThe results indicate that machine learning models can convey meaningful information to support predictions regarding the degree of problem gambling.ConclusionMachine learning models trained using important features showed moderate accuracy in a large-scale Korean adolescent dataset. These findings suggest that the method will help screen adolescents at risk of problem gambling. We believe that expandable machine learning-based approaches will become more powerful as more datasets are collected.

Characterization of Dysregulated lncRNA-Associated ceRNA Network Reveals Novel lncRNAs With ceRNA Activity as Epigenetic Diagnostic Biomarkers for Osteoporosis Risk.

The altered expression of long non-coding RNAs (lncRNAs) has been implicated in the development and human diseases. However, functional roles and regulatory mechanisms of lncRNA as competing endogenous RNAs (ceRNAs) in osteoporosis and their potential clinical implication for osteoporosis risk are largely unexplored. In this study, we performed integrated analysis for paired expression profiles and regulatory relationships of dysregulated lncRNAs, mRNAs, and miRNAs based on “ceRNA hypothesis,” and constructed an osteoporosis-related dysregulated miRNA-mediated lncRNA–mRNA ceRNA network (DysCeNet) composed of 105 nodes (including eight miRNAs, 24 mRNAs, and 73 lncRNAs) and 515 edges. Functional analysis suggested that the DysCeNet was involved in known osteoporosis or bone metabolism-related biological processes and pathways. Then, we performed random forest-based feature selection for 73 lncRNAs with ceRNA activity and identified 25 of 73 lncRNAs as potential diagnostic biomarkers. A random forest-based classifier composed of 25 lncRNA biomarkers (RF-25lncRNA) was developed for predicting osteoporosis risk. Performance evaluation with the leave-one-out cross-validation (LOOCV) procedure showed that the RF-25lncRNA achieved a good performance in distinguishing high- and low-bone mineral density (BMD) subjects in different osteoporosis datasets. Our study for the first time revealed a global view of lncRNA-associated ceRNA regulation in osteoporosis and provided novel lncRNAs with ceRNA activity as candidate epigenetic diagnostic biomarkers for early detection of osteoporosis risk.

Random forest–based feature selection and detection method for drunk driving recognition

A method for drunk driving detection using Feature Selection based on the Random Forest was proposed. First, driving behavior data were collected using a driving simulator at Beijing University of Technology. Second, the features were selected according to the Feature Importance in the random forest. Third, a dummy variable was introduced to encode the geometric characteristics of different roads so that drunk driving under different road conditions can be detected with the same classifier based on the random forest. Finally, the linear discriminant analysis, support vector machine, and AdaBoost classifiers were used and compared with the random forest. The accuracy, F1 score, receiver operating characteristic curve, and area under the curve value were used to evaluate the performance of the classifiers. The results show that Accelerator Depth, Speed, Distance to the Center of the Lane, Acceleration, Engine Revolution, Brake Depth, and Steering Angle have important influences on identifying the drivers’ states and can be used to detect drunk driving. Specifically, the classifiers with Accelerator Depth outperformed the other classifiers without Accelerator Depth. This means that Accelerator Depth is an important feature. Both the AdaBoost and random forest classifiers have an accuracy of 81.48%, which verified the effectiveness of the proposed method.