Sequential Backward Selection Research Articles

Computer-Aided Diagnostic System for Digital Mammography

In this work, Computer-Aided Detection (CADe) and Computer-Aided Diagnosis (CADx) systems are developed and tested using the public and freely available mammographic databases named MIAS and DDSM databases, respectively. CADe system is used to differentiate between normal and abnormal tissues, and it assists radiologists to avoid missing a breast abnormality. At the same time, CADx is developed to distinguish between normal, benign and malignant breast tissues, and it helps radiologists to decide whether a biopsy is needed when reading a diagnostic mammogram or not. Any CAD system is constituted of typical stages including preprocessing and segmentation of mammogram images, extraction of regions of interest (ROI), features removal, features selection and classification. In both proposed CAD systems, ROIs are selected using a window size of 32×32 pixels, then a total of 543 features from four different feature categories are extracted from each ROI and then normalized. After that, the selection of the most relevant features is performed using four different selection methods from MATLAB Pattern Recognition Toolbox v.5 (PRtool5) named Sequential Backward Selection (SBS), Sequential Forward Selection (SFS), Sequential Floating Forward Selection (SFFS) and Branch and Bound Selection (BBS) methods. We also utilized Principal Component Analysis (PCA) as the fifth method to reduce the dimensions of the features set. After that, we used different classifiers such as Support Vector Machines (SVM), K-voting Nearest Neighbor (K-NN), Quadratic Discriminant Analysis (QDA) and Artificial Neural Networks (ANN) for the classification. Both CAD systems have the same implementation stages but different output. CADe systems are designed to detect breast abnormalities while CADx system indicates the likelihood of malignancy of lesions. Finally, we independently compared the performance of all classifiers with each selection method in both modes. The evaluation of the proposed CAD systems is done using performance indices such as sensitivity, specificity, the area under the curve (AUC) of the Receiver Operating Characteristic (ROC) curves, the overall accuracy and Cohen-k factor. Both CAD systems provided encouraging results. These results were different corresponding to the selection method and classifier.

Predicting 30-Day Mortality after an Acute Coronary Syndrome (ACS) using Machine Learning Methods for Feature Selection, Classification and Visualisation

Hybrid combinations of feature selection, classification and visualisation using machine learning (ML) methods have the potential for enhanced understanding and 30-day mortality prediction of patients with cardiovascular disease using population-specific data. Identifying a feature selection method with a classifier algorithm that produces high performance in mortality studies is essential and has not been reported before. Feature selection methods such as Boruta, Random Forest (RF), Elastic Net (EN), Recursive Feature Elimination (RFE), learning vector quantization (LVQ), Genetic Algorithm (GA), Cluster Dendrogram (CD), Support Vector Machine (SVM) and Logistic Regression (LR) were combined with RF, SVM, LR, and EN classifiers for 30-day mortality prediction. ML models were constructed using 302 patients and 54 input variables from the Malaysian National Cardiovascular Disease Database. Validation of the best ML model was performed against Thrombolysis in Myocardial Infarction (TIMI) using an additional dataset of 102 patients. The Self-Organising Feature Map (SOM) was used to visualise mortality-related factors post-ACS. The performance of MLmodels using the area under the curve (AUC) ranged from 0.48 to 0.80. The best-performing model (AUC = 0.80) was a hybrid combination of the RF variable importance method, the sequential backward selection and the RF classifier using five predictors (age, triglyceride, creatinine, troponin, and total cholesterol). Comparison with TIMI using an additional dataset resulted in the best ML model outperforming the TIMI score (AUC = 0.75 vs. AUC = 0.60). The findings of this study will provide a basis for developing an online ML-based population-specific risk scoring calculator.

XGBXSS: An Extreme Gradient Boosting Detection Framework for Cross-Site Scripting Attacks Based on Hybrid Feature Selection Approach and Parameters Optimization

With the widespread popularity of the Internet and the transformation of the world into a global village, Web applications have been drawn increased attention over the years by companies, organizations, and social media, making it a prime target for cyber-attacks. The cross-site scripting attack (XSS) is one of the most severe concerns, which has been highlighted in the forefront of information security experts' reports. In this study, we proposed XGBXSS, a novel web-based XSS attack detection framework based on an ensemble-learning technique using the Extreme Gradient Boosting algorithm (XGboost) with extreme parameters optimization approach. An enhanced feature extraction method is presented to extract the most useful features from the developed dataset. Furthermore, a novel hybrid approach for features selection is proposed, comprising information gain (IG) fusing with sequential backward selection (SBS) to select an optimal subset reducing the computational costs and maintaining the high-performance of detector' simultaneously. The proposed framework has successfully exceeded several tests on the holdout testing dataset and achieved avant-garde results with accuracy, precision, detection probabilities, F-score, false-positive rate, false-negative rate, and AUC-ROC scores of 99.59%, 99.53 %, 99.01%, 99.27%, 0.18%, 0.98%, and 99.41%, respectively. Moreover, it can bridge the existing research gap concerning previous detectors, with a higher detection rate and lesser computational complexity. It also has the potential to be deployed as a self-reliant system, which is efficient enough to defeat such attacks, including zero-day XSS-based attacks.

Improved binary particle swarm optimization for feature selection with new initialization and search space reduction strategies

Feature selection (FS) is an important preprocessing technique for dimensionality reduction in classification problems. Particle swarm optimization (PSO) algorithms have been widely used as the optimizers for FS problems. However, with the increase of data dimensionality, the search space expands dramatically, which proposes significant challenges for optimization methods, including PSO. In this paper, we propose an improved sticky binary PSO (ISBPSO) algorithm for FS. ISBPSO adopts three new mechanisms based on a recently proposed binary PSO variant, sticky binary particle swarm optimization (SBPSO), to improve the evolutionary performance. First, a new initialization strategy using the feature weighting information based on mutual information is proposed. Second, a dynamic bits masking strategy for gradually reducing the search space during the evolutionary process is proposed. Third, based on the framework of memetic algorithms, a refinement procedure conducting genetic operations on the personal best positions of ISBPSO is used to alleviate the premature convergence problem. The results on 12 UCI datasets show that ISBPSO outperforms six benchmark PSO-based FS methods and two conventional FS methods (sequential forward selection and sequential backward selection) — ISBPSO obtains either higher or similar accuracies with fewer features in most cases. Moreover, ISBPSO substantially reduces the computation time compared with benchmark PSO-based FS methods. Further analysis shows that all the three proposed mechanisms are effective for improving the search performance of ISBPSO.

Genetic-based Pruning Technique for Ant-Miner Classification Algorithm

Ant colony optimization (ACO) is a well-known algorithm from swarm intelligence that plays an essential role in obtaining rich solutions to complex problems with wide search space. ACO is successfully applied to different application problems involving rules-based classification through an ant-miner classifier. However, in the ant-miner classifier, rule-pruning suffers from the problem of nesting effect origins from the method of greedy Sequential Backward Selection (SBS) in term selection, thereby depriving the opportunity of obtaining a good pruned rule by adding/removing the terms during the pruning process. This paper presents an extension to the Ant-Miner, namely the genetic algorithm Ant-Miner (GA-Ant Miner), which incorporates the use of GA as a key aspect in the design and implementation of a new rule pruning technique. This pruning technique consists of three fundamental procedures: an initial population Ant-Miner, crossover to prune the rule, and mutation to diversify the pruned classification rule. The GA-Ant Miner performance is tested and compared with the most related ant-mining classifiers, including the original Ant-Miner, ACO/ PSO2, TACO-Miner, CAnt-Miner, and Ant-Miner with a hybrid pruner, across various public available UCI datasets. These datasets are varied in terms of instance number, feature size, class number, and the application domains. Overall, the performance results indicate that the GA-Ant Miner classifier outperforms the other five classifiers in the classification accuracy and model size. Furthermore, the experimental results using statistical test prove that GA-Ant Miner is the best classifier when considering the multi objectives (i.e., accuracy and model size ranks).

The crash injury severity prediction of traffic accident using an improved wrappers feature selection algorithm

The crash injury severity prediction of traffic accident is the prerequisite for driving decisions under dangerous conditions. This paper puts forward a new wrapper feature selection algorithm to improve the prediction accuracy of the injury severity. Firstly, based on Gini index and Mutual information, the importance of factors is analyzed by the comprehensive evaluation criteria, according to the importance score, the factors were selected by sequential backward selection. Secondly, three models, that is RF (Random Forest), C4.5 (C4.5 Decision Tree) and SVM (Support Vector Machine) are established to verify the prediction accuracy of selected subsets. In order to verify the proposed algorithm, based on the FARS(Fatality Analysis Reporting System) in the NHTSA(National Highway Traffic Safety Administration), a total of 6,295 rear-end accidents including 23 candidate factors were screened, the results demonstrate that the accuracy of the three models increases with the elimination of redundant variables, among them the SVM model has better predictive performance, the significant factors are obtained which include seat belts, airbag, the relative speed, vehicle weight, etc. finally the sensitivity analysis of significant factors is implemented and some suggestions are given to improve traffic safety.

Diagnosing Parkinson’s disease by means of ensemble classification of patients’ voice samples

This paper proposes a method for diagnosing Parkinson’s disease using ensemble classification of patient voice samples. Conducted research concerned testing the parameters of the ensemble of classifiers, in terms of types and numbers of classifiers included in it. More than a dozen popular classifiers were considered in the study. Additionally, for each of the tested classifiers, a set of features of voice samples were selected, for which a given classifier showed the highest classification efficiency. The Sequential Backward Selection (SBS), which belongs to the wrapper methods, was used for feature selection. The ensemble of classifiers was then tested in two cases, with all features considered and including only those indicated by the SBS method. The obtained results were compared with each other.All experiments were performed on a publicly available database containing voice samples of Parkinson’s patients and healthy patients. This database can be found in the UCI archives.

Predicting Critical Bicycle-Vehicle Conflicts at Signalized Intersections

Continuous development of urban infrastructure with a focus on sustainable transportation has led to a proliferation of vulnerable road users (VRUs), such as bicyclists and pedestrians, at intersections. Intersection safety evaluation has primarily relied on historical crash data. However, due to several limitations, including rarity, unpredictability, and irregularity of crash occurrences, quantitative and qualitative analyses of crashes may not be accurate. To transcend these limitations, intersection safety can be proactively evaluated by quantifying near-crashes using alternative measures known as surrogate safety measures (SSMs). This study focuses on developing models to predict critical near-crashes between vehicles and bicycles at intersections based on SSMs and kinematic data. Video data from ten signalized intersections in the city of San Diego were employed to train logistic regression (LR), support vector machine (SVM), and random forest (RF) models. A variation of time-to-collision called T2 and postencroachment time (PET) were used to specify monitoring periods and to identify critical near-crashes, respectively. Four scenarios were created using two thresholds of 5 and 3 s for both PET and T2. In each scenario, five monitoring period lengths were examined. The RF model was superior compared to other models in all different scenarios and across different monitoring period lengths. The results also showed a small trade-off between model performance and monitoring period length, identifying models with monitoring period lengths of 10 and 20 frames performed slightly better than those with lower or higher lengths. Sequential backward and forward feature selection methods were also applied that enhanced model performance. The best RF model had recall values of 85% or higher across all scenarios. Also, RF prediction models performed better when considering just the rear-end near-crashes with recalls of above 90%.

A portable HCI system‐oriented EEG feature extraction and channel selection for emotion recognition

Emotion recognition has become an important component of human–computer interaction systems. Research on emotion recognition based on electroencephalogram (EEG) signals are mostly conducted by the analysis of all channels' EEG signals. Although some progresses are achieved, there are still several challenges such as high dimensions, correlation between different features and feature redundancy in the realistic experimental process. These challenges have hindered the applications of emotion recognition to portable human–computer interaction systems (or devices). This paper explores how to find out the most effective EEG features and channels for emotion recognition so as to only collect data as less as possible. First, discriminative features of EEG signals from different dimensionalities are extracted for emotion classification, including the first difference, multiscale permutation entropy, Higuchi fractal dimension, and discrete wavelet transform. Second, relief algorithm and floating generalized sequential backward selection algorithm are integrated as a novel channel selection method. Then, support vector machine is employed to classify the emotions for verifying the performance of the channel selection method and extracted features. At last, experimental results demonstrate that the optimal channel set, which are mostly located at the frontal, has extremely high similarity on the self-collected data set and the public data set and the average classification accuracy is achieved up to 91.31% with the selected 10-channel EEG signals. The findings are valuable for the practical EEG-based emotion recognition systems.

Pattern Recognition of Cognitive Load Using EEG and ECG Signals.

The matching of cognitive load and working memory is the key for effective learning, and cognitive effort in the learning process has nervous responses which can be quantified in various physiological parameters. Therefore, it is meaningful to explore automatic cognitive load pattern recognition by using physiological measures. Firstly, this work extracted 33 commonly used physiological features to quantify autonomic and central nervous activities. Secondly, we selected a critical feature subset for cognitive load recognition by sequential backward selection and particle swarm optimization algorithms. Finally, pattern recognition models of cognitive load conditions were constructed by a performance comparison of several classifiers. We grouped the samples in an open dataset to form two binary classification problems: (1) cognitive load state vs. baseline state; (2) cognitive load mismatching state vs. cognitive load matching state. The decision tree classifier obtained 96.3% accuracy for the cognitive load vs. baseline classification, and the support vector machine obtained 97.2% accuracy for the cognitive load mismatching vs. cognitive load matching classification. The cognitive load and baseline states are distinguishable in the level of active state of mind and three activity features of the autonomic nervous system. The cognitive load mismatching and matching states are distinguishable in the level of active state of mind and two activity features of the autonomic nervous system.

Computationally identifying hot spots in protein-DNA binding interfaces using an ensemble approach

BackgroundProtein-DNA interaction governs a large number of cellular processes, and it can be altered by a small fraction of interface residues, i.e., the so-called hot spots, which account for most of the interface binding free energy. Accurate prediction of hot spots is critical to understand the principle of protein-DNA interactions. There are already some computational methods that can accurately and efficiently predict a large number of hot residues. However, the insufficiency of experimentally validated hot-spot residues in protein-DNA complexes and the low diversity of the employed features limit the performance of existing methods.ResultsHere, we report a new computational method for effectively predicting hot spots in protein-DNA binding interfaces. This method, called PreHots (the abbreviation of Predicting Hotspots), adopts an ensemble stacking classifier that integrates different machine learning classifiers to generate a robust model with 19 features selected by a sequential backward feature selection algorithm. To this end, we constructed two new and reliable datasets (one benchmark for model training and one independent dataset for validation), which totally consist of 123 hot spots and 137 non-hot spots from 89 protein-DNA complexes. The data were manually collected from the literature and existing databases with a strict process of redundancy removal. Our method achieves a sensitivity of 0.813 and an AUC score of 0.868 in 10-fold cross-validation on the benchmark dataset, and a sensitivity of 0.818 and an AUC score of 0.820 on the independent test dataset. The results show that our approach outperforms the existing ones.ConclusionsPreHots, which is based on stack ensemble of boosting algorithms, can reliably predict hot spots at the protein-DNA binding interface on a large scale. Compared with the existing methods, PreHots can achieve better prediction performance. Both the webserver of PreHots and the datasets are freely available at: http://dmb.tongji.edu.cn/tools/PreHots/.

An Improved Hybrid Feature Selection Algorithm for Electric Charge Recovery Risk

In order to extract more information that affects customer arrears behavior, the feature extraction method is used to extend the low-dimensional features to the high-dimensional features for the warning problem of user arrears risk model of electric charge recovery (ECR). However, there are many irrelevant or redundant features in data, which affect prediction accuracy. In order to reduce the dimension of the feature and improve the prediction result, an improved hybrid feature selection algorithm is proposed, integrating nonlinear inertia weight binary particle swarm optimization with shrinking encircling and exploration mechanism (NBPSOSEE) with sequential backward selection (SBS), namely, NBPSOSEE-SBS, for selecting the optimal feature subset. NBPSOSEE-SBS can not only effectively reduce the redundant or irrelevant features from the feature subset selected by NBPSOSEE but also improve the accuracy of classification. The experimental results show that the proposed NBPSOSEE-SBS can effectively reduce a large number of redundant features and stably improve the prediction results in the case of low execution time, compared with one state-of-the-art optimization algorithm, and seven well-known wrapper-based feature selection approaches for the risk prediction of ECR for power customers.

Analysis of autonomic nervous pattern in hypertension based on short-term heart rate variability.

Physiological studies have found that the autonomic nervous system plays an important role in controlling blood pressure values. This paper, based on machine learning approaches, analysed short-term heart rate variability to determine differences in autonomic nervous function between hypertensive patients and normal population. The electrocardiogram (ECG) of hypertensive patients are 137 ECG recordings provided by Smart Health for Assessing the Risk of Events via ECG (SHAREE database). The RR intervals of healthy subjects include the data of 18 subjects from the MIT-BIH Normal Sinus Rhythm Database (nsrdb) and 54 subjects from the Normal Sinus Rhythm RR Interval Database (nsr2db). In this paper, each RR segment includes continuous 500 beats. Seventeen features were extracted to distinguish the hypertensive heart beat rhythms from the normal ones, and Kolmogorov-Smirnov test and sequential backward selection (SBS) were applied to get the best feature combinations. In addition, support vector machine (SVM), k-nearest neighbor (KNN) and random forest (RF) were applied as classifiers in the study. The performance of each classifier was evaluated independently using the leave-one-subject-out validation method. The best predictive model was based on RF and enabled to identify hypertensive patients by five features with an accuracy of 86.44%. The best five HRV features are sample entropy (SampEn), very low frequency spectral powers (VLF), root mean square of successful differences (RMSSD), ratio of low frequency spectral powers and high frequency spectral powers (LF/HF) and vector angle index (VAI). The results of the study show sympathetic overactivity and decreased parasympathetic tone in hypertensive patients.

Dominant point detection based on suboptimal feature selection methods

This paper presents a viable alternative solution for dominant point detection predicated on the comparison of suboptimal feature selection methods. Suboptimal feature selection methods are utilized as standard criteria to identify dominant points. Considering that all of the combinations of points comprise many sets, an algorithm that eliminates some of them is affirmed and illustrated. The sequential backward selection, sequential forward selection, generalized sequential forward selection, generalized sequential backward selection and plus l-take away r selection methods are performed on the remaining points to extract the dominant points. The simulation results exhibit that this method is significantly more effective and efficient in comparison to other proposed methods.

Sequential Feature Selection in Customer Churn Prediction Based on Naive Bayes

The customer churn prediction is an important business strategy for the company. The ease of switching operators is one of the serious challenges that must be faced by the telecommunications industry. To get new customers requires a much higher cost than maintaining existing customers. Customer churn refers to the periodic loss of customers in an organization. To retain existing customers, organizations must improve customer service, improve product quality, and must be able to know in advance which customers have the possibility of leaving the organization. By predicting customer churn, companies can immediately take action to retain customers. Prediction can be done by analysing customer data using data mining techniques. This study proposes to implement feature selection to select relevant features and can provide improved performance in customer churn prediction models. Some proposed feature selection techniques are Sequential Forward Selection (SFS), Sequential Backward Selection (SBS), Sequential Forward Floating Selection (SFFS), Sequential Forward Floating Selection (SBFS), Sequential Backward Floating Selection (SBFS). The classification algorithm used to classify is Naive Bayes. The model that provides the best performance value is the model that implements Sequential Backward Selection (SBS) and Sequential Backward Floating Selection (SBFS) feature selection technique with feature number 19.

Leaf Area Index Estimation Algorithm for GF-5 Hyperspectral Data Based on Different Feature Selection and Machine Learning Methods

Leaf area index (LAI) is an essential vegetation parameter that represents the light energy utilization and vegetation canopy structure. As the only in-operation hyperspectral satellite launched by China, GF-5 is potentially useful for accurate LAI estimation. However, there is no research focus on evaluating GF-5 data for LAI estimation. Hyperspectral remote sensing data contains abundant information about the reflective characteristics of vegetation canopies, but these abound data also easily result in a dimensionality curse. Therefore, feature selection (FS) is necessary to reduce data redundancy to achieve more reliable estimations. Currently, machine learning (ML) algorithms have been widely used for FS. Moreover, the same ML algorithm is usually conducted for both FS and regression in LAI estimation. However, no evidence suggests that this is the optimal solution. Therefore, this study focuses on evaluating the capacity of GF-5 spectral reflectance for estimating LAI and the performances of different combination of FS and ML algorithms. Firstly, the PROSAIL model, which coupled leaf optical properties model PROSPECT and the scattering by arbitrarily inclined leaves (SAIL) model, was used to generate simulated GF-5 reflectance data under different vegetation and soil conditions, and then three FS methods, including random forest (RF), K-means clustering (K-means) and mean impact value (MIV), and three ML algorithms, including random forest regression (RFR), back propagation neural network (BPNN) and K-nearest neighbor (KNN) were used to develop nine LAI estimation models. The FS process was conducted twice using different strategies: Firstly, three FS methods were conducted to search the lowest dimension number, which maintained the estimation accuracy of all bands. Then, the sequential backward selection (SBS) method was used to eliminate the bands having minimal impact on LAI estimation accuracy. Finally, three best estimation models were selected and evaluated using reference LAI. The results showed that although the RF_RFR model (RF used for feature selection and RFR used for regression) achieved reliable LAI estimates (coefficient of determination (R2) = 0.828, root mean square error (RMSE) = 0.839), the poor performance (R2 = 0.763, RMSE = 0.987) of the MIV_BPNN model (MIV used for feature selection and BPNN used for regression) suggested using feature selection and regression conducted by the same ML algorithm could not always ensure an optimal estimation. Moreover, RF selection preserved the most informative bands for LAI estimation so that each ML regression method could achieve satisfactory estimation results. Finally, the results indicated that the RF_KNN model (RF used as feature selection and KNN used for regression) with seven GF-5 spectral band reflectance achieved the better estimation results than others when validated by simulated data (R2 = 0.834, RMSE = 0.824) and actual reference LAI (R2 = 0.659, RMSE = 0.697).

Feature Selection Based on Naive Bayes for Caesarean Section Prediction

Data mining using machine learning algorithms can be used to help analyze historical data to predict the need for a caesarean section. The dataset used for predicting caesarean section has many features, but those features have the possibility of redundancy or irrelevance that can cause a decrease in classifier performance. This research proposes a model that implements feature selection to select relevant features and can provide improved performance predictions for caesarean section. Some proposed feature selection techniques are Sequential Forward Selection (SFS), Sequential Backward Selection (SBS), Sequential Forward Floating Selection (SFFS), Sequential Forward Floating Selection (SBFS), Sequential Backward Floating Selection (SBFS), and selectKBest. The classification algorithm used to classify is Naive Bayes. The model that gives the best performance value is the model that applies the SelectKbest as feature selection.

Accuracy comparison of dimensionality reduction techniques to determine significant features from IMU sensor-based data to diagnose vestibular system disorders

This study is a significant step gone to develop Machine Learning (ML) algorithm to apply to gait sensory information collected from people to identify Vestibular System (VS) disorders. Although ML is widely used as diagnostic tool in medical decision-making, there is not much research done on application of ML methods to identify VS imperfections. In this paper, we compared the accuracies of two dimensionality-reduction techniques to use with SVM with Gaussian Kernel: Feature Selection (FS) and Feature Transformation (FT) methods. T-test and Sequential Backward Selection (SBS) were used for FS and Principal Component Analysis (PCA) and Kernel Principal Component Analysis (KPCA) with polynomial and Gaussian kernels were used as FT method. Both methods were applied to the dataset formed by 22 features collected from 37 people, of whom 21 were healthy and 16 subjects had VS-disorders. The highest accuracy among FT methods was 89.2%, while it was 81.1% for FS method. SVM with Gaussian Kernel, trained with the dataset of reduced dimensionality, had computation time of few hundreds of milliseconds, which makes real-time data processing possible. The importance of this work will obviously increase with the increase in the number of initial features. As a next step, we aim to increase dataset and use additional features extracted from pressure sensors placed under the feet. We also aim to use time domain characteristics of the features to increase overall accuracy as a next step.

A Complex Fault Diagnostic Approach of Active Distribution Network Based on SBS-SFS Optimized Multi-SVM

After renewable energy distributed generator (DG) is connected to the power grid, traditional diverse-electric-information-based fault diagnosis approaches are not suitable for an active distributed network (ADN) due to the weak characteristics of fault current. Thus, this paper proposes a comprehensive nonformula fault diagnostic approach of ADN using only voltage as input. In the preprocess, sequential forward selection (SFS) and sequential backward selection (SBS) are utilized to optimize the input feature matrix of the sample in order to reduce the information redundancy of multiple measuring points in ADN. Then, a single “1-a-1” support vector machine (SVM) classifier is used for fault identification, and a multi-SVM, with radial basis function (RBF) as the kernel function, is applied to identify the location and fault type. To prove the proposed method is adaptable for ADN, two direct drive fans are used as a DG to test the IEEE 33 node model at every 10% of the line under three operating conditions that include all cases of distributed power generation in ADN. Results comparing real-time and historical data show that the proposed multi-SVM model reaches an average fault type diagnosis accuracy of 97.27%, with a fault identification accuracy of 96%. A backpropagation neural network is then compared to the proposed model. The results show the superior performance of the SBS-SFS optimized multi-SVM. This model can be usefully applied to the fault diagnosis of new energy sources with distributed power access to distribution networks.

FastVGBS: A Fast Version of the Volume-Gradient-Based Band Selection Method for Hyperspectral Imagery

Recently, the volume-gradient-based band selection (VGBS) method has attracted more and more attention in the field of band selection. It is a ranking-based unsupervised algorithm which applies the sequential backward selection strategy to successively remove the most abundant band. The key finding of VGBS is that the band redundancy corresponds to the volume gradient matrix with respect to hyperspectral images. However, we have found that VGBS requires to update the gradient matrix after each band removal, which includes the calculation of the matrix inverse, determinant, and multiplication, and thus is time-consuming when the number of bands is large. In this letter, we first find that the norm of the row of the gradient matrix has a one-to-one correspondence to the diagonal element of the covariance matrix of the image. Further, we develop a recursive formula to calculate the inverse of the covariance matrix. The experimental results show the effectiveness of the method, i.e., we can reduce the computational complexity of VGBS with an order of magnitude.