Improved Feature Selection Algorithm Research Articles

Prediction of SO2 concentration at desulfurization outlet of thermal power units based on ReliefF-SC and ISAO-ARELM

Abstract The flue gas desulfurization (FGD) system of thermal power units operates under complex conditions and exhibits significant nonlinearity. Establishing an accurate prediction model for outlet SO2 concentration is crucial for optimizing the control of the FGD system. The study constructs an autoregressive limit learning machine (ARELM) prediction model for SO2 concentration at the desulfurization outlet of thermal power units, leveraging the improved feature selection algorithm ReliefF-SC and the improved snow ablation optimizer (ISAO). Initially, the time delays of the input variables are compensated and the ReliefF-SC algorithm, which incorporates the Spearman correlation coefficient and cosine similarity, is designed to obtain the optimal feature set for predicting SO2 concentration at the outlet. To enhance the ELM's capacity to process time-series data, the autoregressive (AR) concept is integrated into the ELM framework. Furthermore, to mitigate the impact of random initialization of ARELM network parameters on model stability, the ISAO algorithm is proposed by introducing the sine-cosine position update strategy and adaptive adjustment of subpopulation size. Finally, experimental validation is conducted using actual plant operation data. The results indicate that the established SO2 concentration prediction model for desulfurization outlets of thermal power units is highly accurate and offers valuable theoretical guidance and technical support for optimizing the control of the FGD systems.

SVM-RLF-DNN: A DNN with reliefF and SVM for automatic identification of COVID from chest X-ray and CT images.

To develop an advanced determination technology for detecting COVID-19 patterns from chest X-ray and CT-scan films with distinct applications of deep learning and machine learning methods. The newly enhanced proposed hybrid classification network (SVM-RLF-DNN) comprises of three phases: feature extraction, selection and classification. The in-depth features are extracted from a series of 3×3 convolution, 2×2 max polling operations followed by a flattened and fully connected layer of the deep neural network (DNN). ReLU activation function and Adam optimizer are used in the model. The ReliefF is an improved feature selection algorithm of Relief that uses Manhattan distance instead of Euclidean distance. Based on the significance of the feature, the ReliefF assigns weight to each extracted feature received from a fully connected layer. The weight to each feature is the average of k closest hits and misses in each class for a neighbouring instance pair in multiclass problems. The ReliefF eliminates lower-weight features by setting the node value to zero. The higher weights of the features are kept to obtain the feature selection. At the last layer of the neural network, the multiclass Support Vector Machine (SVM) is used to classify the patterns of COVID-19, viral pneumonia and healthy cases. The three classes with three binary SVM classifiers use linear kernel function for each binary SVM following a one-versus-all approach. The hinge loss function and L2-norm regularization are selected for more stable results. The proposed method is assessed on publicly available chest X-ray and CT-scan image databases from Kaggle and GitHub. The performance of the proposed classification model has comparable training, validation, and test accuracy, as well as sensitivity, specificity, and confusion matrix for quantitative evaluation on five-fold cross-validation. Our proposed network has achieved test accuracy of 98.48% and 95.34% on 2-class X-rays and CT. More importantly, the proposed model's test accuracy, sensitivity, and specificity are 87.9%, 86.32%, and 90.25% for 3-class classification (COVID-19, Pneumonia, Normal) on chest X-rays. The proposed model provides the test accuracy, sensitivity, and specificity of 95.34%, 94.12%, and 96.15% for 2-class classification (COVID-19, Non-COVID) on chest CT. Our proposed classification network experimental results indicate competitiveness with existing neural networks. The proposed neural network assists clinicians in determining and surveilling the disease.

Improved Feature Selection Algorithm for Intrusion Detection Using Data Mining Approach

With the rapid growth of Internet applications, there are more and more intrusions into network systems. In this case, it is necessary to provide security for the network through effective intrusion detection and prevention methods. This can mainly be achieved by creating effective interruption detection systems using efficient algorithms that can identify abnormal activities in network traffic and safeguard network resources from unlawful attack by interlopers. Although many interruption recognition frameworks have been proposed before, existing network intrusion detection has limitations in terms of accuracy and detection time. To overcome these shortcomings, In this paper, we propose a new intrusion detection system by developing a new intelligent feature selection algorithm based on conditional random fields (CRF) to optimize the number of features. Furthermore, algorithms based on existing hierarchical methods (LA) In this paper, we propose another interrupt recognition framework, fostering a book. Compared with the existing methods, the interruption identification framework provides high precision and achieves the efficiency of attack detection. The main advantages of this system are reduced detection time, improved classification accuracy and lower false positive rate.

A network traffic identification method based on AutoEncoder - a feature selection algorithm

Traffic identification methods consider a large number of traffic features, resulting in low identification efficiency. To address the efficiency problem of traffic recognition, this paper proposes an efficient network traffic recognition method, AutoEncoder-based traffic recognition (AE-NTI). The method first preprocesses the original dataset and converts it into a two-dimensional grayscale image. Then, feature selection is performed by an improved feature selection algorithm based on AutoEncoder. The algorithm consists of a feature scorer, which globally scores all features, and a feature selector, which selects the highest scoring features to reconstruct the original data. Finally, the convolutional neural network structure is adjusted so that the network traffic can be identified. The experimental results show that the method has significantly improved in recognition accuracy and model fitting speed.

Machine Learning Prediction of CO Adsorption Energies and Properties of Layered Alloys Using an Improved Feature Selection Algorithm

Layered alloys are widely studied as designable catalysts. As a surface probe molecule, CO adsorption energy is not only employed to characterize surface properties but also used as the catalytic activity descriptor in various reactions. With the aid of high-throughput computing technology, we calculated CO adsorption energies on 3729layered alloy surfaces. To obtain CO adsorption energies, the d-band center and d-band skewness, and the stability of all the remaining layered alloys (8415) of 23 transition metals, we collected 91 features that do not require time-consuming quantum chemistry calculations (non-QC features) and 40 features from quantum chemistry calculations (QC features). To reduce the feature dimension and overcome overfitting problems, we proposed a modified sequential feature selection (SFS) wrapper method to identify (sub)optimal subsets. Two supervised light gradient boosting machine regression (LGBMR) machine learning (ML) regression models were established using the identified subsets. It is demonstrated that the size of the feature subset converges rapidly, and the performance of the model with size nine is already quite satisfactory. The ML model of the non-QC features outperforms that of QC features. Using the ML models established with non-QC features, we predicted the CO adsorption energies and the electronic structure–properties (d-band center, d-band skewness) and stability of 8415 layered alloys. Based on the four conditions (CO adsorption energy, stability, price, and surface segregation), potential alloy catalysts for CO2 to methanol were screened out of 12144 layered alloys.

Improved Multi-Verse Optimizer Feature Selection Technique With Application To Phishing, Spam, and Denial Of Service Attacks

Intelligent classification systems proved their merits in different fields including cybersecurity. However, most cybercrime issues are characterized of being dynamic and not static classification problems where the set of discriminative features keep changing with time. This indeed requires revising the cybercrime classification system and pick a group of features that preserve or enhance its performance. Not only this but also the system compactness is regarded as an important factor to judge on the capability of any classification system where cybercrime classification systems are not an exception. The current research proposes an improved feature selection algorithm that is inspired from the well-known multi-verse optimizer (MVO) algorithm. Such an algorithm is then applied to 3 different cybercrime classification problems namely phishing websites, spam, and denial of service attacks. MVO is a population-based approach which stimulates a well-known theory in physics namely multi-verse theory. MVO uses the black and white holes principles for exploration, and wormholes principle for exploitation. A roulette selection schema is used for scientifically modeling the principles of white hole and black hole in exploration phase, which bias to the good solutions, in this case the solutions will be moved toward the best solution and probably to lose the diversity, other solutions may contain important information but didn’t get chance to be improved. Thus, this research will improve the exploration of the MVO by introducing the adaptive neighborhood search operations in updating the MVO solutions. The classification phase has been done using a classifier to evaluate the results and to validate the selected features. Empirical outcomes confirmed that the improved MVO (IMVO) algorithm is capable to enhance the search capability of MVO, and outperform other algorithm involved in comparison.

LBP-based bird sound classification using improved feature selection algorithm

LBP-based bird sound classification using improved feature selection algorithm

Improved Multi-Verse Optimizer Feature Selection Technique With Application To Phishing, Spam, and Denial Of Service Attacks

Intelligent classification systems proved their merits in different fields including cybersecurity. However, most cybercrime issues are characterized of being dynamic and not static classification problems where the set of discriminative features keep changing with time. This indeed requires revising the cybercrime classification system and pick a group of features that preserve or enhance its performance. Not only this but also the system compactness is regarded as an important factor to judge on the capability of any classification system where cybercrime classification systems are not an exception. The current research proposes an improved feature selection algorithm that is inspired from the well-known multi-verse optimizer (MVO) algorithm. Such an algorithm is then applied to 3 different cybercrime classification problems namely phishing websites, spam, and denial of service attacks. MVO is a population-based approach which stimulates a well-known theory in physics namely multi-verse theory. MVO uses the black and white holes principles for exploration, and wormholes principle for exploitation. A roulette selection schema is used for scientifically modeling the principles of white hole and black hole in exploration phase, which bias to the good solutions, in this case the solutions will be moved toward the best solution and probably to lose the diversity, other solutions may contain important information but didn’t get chance to be improved. Thus, this research will improve the exploration of the MVO by introducing the adaptive neighborhood search operations in updating the MVO solutions. The classification phase has been done using a classifier to evaluate the results and to validate the selected features. Empirical outcomes confirmed that the improved MVO (IMVO) algorithm is capable to enhance the search capability of MVO, and outperform other algorithm involved in comparison.

An improve feature selection algorithm for defect detection of glass bottles

Defect detection is an effective technology to guarantee the quality of the products. In this paper, the defect detection of glass bottles is taken as a classification problem. Features abstracted from the knocking sound signals of the glass bottles are used as the input of the classifier. Because the detection of bottles usually should be done in real time, feature selection which could reduce the dimensionality of data and simplify the learned model has become an important technology in this work. Based on our previous proposed algorithm named Shuffled Frog Leaping Algorithm-Improved Minimal Redundancy Maximal Relevance(SFLA-ImRMR), this paper focuses on improving the classification performance and proposes an improved feature selection algorithm. In the proposed algorithm, feature selection is combined with the training of the classifier, the wrapper approach is used to evaluate the selected features, Shuffled Frog Leaping Algorithm(SFLA) is used as the search algorithm, and BP Neural Network(BPNN) is used as the classifier. The proposed algorithm is named SFLA-ImRMR-BP and tested in the data sets built by the knocking sound signals of the glass bottles. Compared with SFLA-ImRMR and some other feature selection algorithms based on Evolutionary Computation(EC), features selected by SFLA-ImRMR-BP can achieve the highest classification performance.

Learning robust affinity graph representation for multi-view clustering

Recently, an increasingly pervasive trend in real-word applications is that the data are collected from multiple sources or represented by multiple views. Owing to the powerful ability of affinity graph in capturing the structural relationships among samples, constructing a robust and meaningful affinity graph has been extensively studied, especially in spectral clustering tasks. However, conventional spectral clustering extended to multi-view scenarios cannot obtain the satisfactory performance due to the presence of noise and the heterogeneity among different views. In this paper, we propose a robust affinity graph learning framework to deal with this issue. First, we employ an improved feature selection algorithm that integrates the advantages of hypergraph embedding and sparse regression to select significant features such that more robust graph Laplacian matrices for various views on this basis can be constructed. Second, we model hypergraph Laplacians as points on a Grassmann manifold and propose a Consistent Affinity Graph Learning (CAGL) algorithm to fuse all views. CAGL aims to learn a latent common affinity matrix shared by all Laplacian matrices by taking both the clustering quality evaluation criterion and the view consistency loss into account. We also develop an alternating descent algorithm to optimize the objective function of CAGL. Experiments on five publicly available datasets demonstrate that our proposed method obtains promising results compared with state-of-the-art methods.

Enhancing BCI-Based Emotion Recognition Using an Improved Particle Swarm Optimization for Feature Selection.

Electroencephalogram (EEG) signals have been widely used in emotion recognition. However, the current EEG-based emotion recognition has low accuracy of emotion classification, and its real-time application is limited. In order to address these issues, in this paper, we proposed an improved feature selection algorithm to recognize subjects’ emotion states based on EEG signal, and combined this feature selection method to design an online emotion recognition brain-computer interface (BCI) system. Specifically, first, different dimensional features from the time-domain, frequency domain, and time-frequency domain were extracted. Then, a modified particle swarm optimization (PSO) method with multi-stage linearly-decreasing inertia weight (MLDW) was purposed for feature selection. The MLDW algorithm can be used to easily refine the process of decreasing the inertia weight. Finally, the emotion types were classified by the support vector machine classifier. We extracted different features from the EEG data in the DEAP data set collected by 32 subjects to perform two offline experiments. Our results showed that the average accuracy of four-class emotion recognition reached 76.67%. Compared with the latest benchmark, our proposed MLDW-PSO feature selection improves the accuracy of EEG-based emotion recognition. To further validate the efficiency of the MLDW-PSO feature selection method, we developed an online two-class emotion recognition system evoked by Chinese videos, which achieved good performance for 10 healthy subjects with an average accuracy of 89.5%. The effectiveness of our method was thus demonstrated.

A study on Subtractive Pixel Adjacency Matrix features

Subtractive Pixel Adjacency Matrix (SPAM) features perform well in detecting spatial-domain steganographic algorithm. Further, some methods of SPAM features can be applied to rich models and steganalysis based on deep learning. Therefore, this paper presents a study on SPAM features and it is divided into two parts: in the first part, impact of spatial-domain steganographic on difference between adjacent pixels is first analyzed. Then, three SPAM features are proposed with the same range of differences and different orders of Markov chain. Following that, the influences of order of Markov chain and range of differences on SPAM features are analyzed, and we find that detection accuracy of SPAM features increases with the range of differences increasing; in the second part, SPAM feature is first divided into several modules according to the conclusion. Then, taking detection accuracy of support vector machine (SVM) classifier and mutual information as metrics and module as a unit, a Novel Feature Selection (NFS) algorithm and an Improved Feature Selection algorithm are proposed. Experimental results show that the NFS algorithm can achieve higher detection accuracy than several existing algorithms.

Accessible Melanoma Detection Using Smartphones and Mobile Image Analysis

We investigate the design of an entire mobile imaging system for early detection of melanoma. Different from previous work, we focus on smartphone-captured visible light images. Our design addresses two major challenges. First, images acquired using a smartphone under loosely-controlled environmental conditions may be subject to various distortions, and this makes melanoma detection more difficult. Second, processing performed on a smartphone is subject to stringent computation and memory constraints. In our work, we propose a detection system that is optimized to run entirely on the resource-constrained smartphone. Our system intends to localize the skin lesion by combining a lightweight method for skin detection with a hierarchical segmentation approach using two fast segmentation methods. Moreover, we study an extensive set of image features and propose new numerical features to characterize a skin lesion. Furthermore, we propose an improved feature selection algorithm to determine a small set of discriminative features used by the final lightweight system. In addition, we study the human–computer interface (HCI) design to understand the usability and acceptance issues of the proposed system. Our extensive evaluation on an image dataset provided by National Skin Center - Singapore (117 benign nevi and 67 malignant melanoma) confirms the effectiveness of the proposed system for melanoma detection: 89.09% sensitivity at specificity $\geq$ 90%.

An improved feature selection algorithm based on graph clustering and ant colony optimization

Dimensionality reduction is an important preprocessing step to improve the performance of machine learning algorithms. Feature selection methods can efficiently speed up the learning process and improve the overall classification accuracy by reducing the computational complexity. Among the feature selection methods, multivariate methods are more effective in removing irrelevant and redundant features. An efficient multivariate feature selection method, optimization method, called ‘graph clustering based ant colony optimization (GCACO)’ has been recently introduced and shown to outperform other well-known feature selection methods. In the GCACO, features are divided into communities (clusters) in the entire feature space represented as a graph by an efficient community detection algorithm. An ACO-based search strategy is then used to select an optimal feature subset from the initial set of features. In this paper, a modified GCACO algorithm called MGCACO is presented to significantly improve the performance of the GCACO. Performance of the MGCACO algorithm was assessed by testing it on several standard benchmark datasets and sleep EEG data. The performance of the MGCACO was compared to those obtained using the original GCACO and other well-known filtering methods available in the literature. The MGCACO achieved superior performance over the GCACO and other univariate and multivariate algorithms with up to 10%. The MGCACO also exhibited higher efficiency in reducing the number of features all by keeping the classification accuracy maximum.

An Improved Feature Selection Algorithm Based on Ant Colony Optimization

The diversity and complexity of network data bring great challenges to data classification technology. Feature selection has always been an important and difficult problem in classification technology. To improve the classification performance of the classifier, an improved feature selection algorithm, FACO, is proposed by combining the ant colony optimization algorithm and feature selection. A fitness function is designed, and the pheromone updating rule is optimized to effectively eliminate redundant features and prevent feature selection from falling into a local optimum. The experimental results show that the classification accuracy of the classifier can be significantly improved by selecting the data features using the FACO algorithm, which is of practical significance.

An Improved Feature Selection Algorithm for Ordinal Classification

An Improved Feature Selection Algorithm for Ordinal Classification

An Improved Immune Clone Selection Algorithm for Palm Bio-impedance Spectroscopy

A kind of effective feature model of palm BIS data is presented according to the features of Palm bio-impedance spectroscopy (BIS) data. Based on immune clone algorithm and least squares method, an improved palm BIS feature selection algorithm is established,which can be applied to obtain the optimal feature subset that can be completely represented the palm BIS data, Finally, the algorithm is compared with other algorithms. The experimental results show that the accuracy of the feature subset obtained by the algorithm has reached 93.2 in SVM classification algorithm test.Therefore, the algorithm in this article is valid and reliable , which is of high theoretical and practical value. Full Text: PDF DOI: http://dx.doi.org/10.11591/telkomnika.v16i2.8965

Selecting Key Features for Remote Sensing Classification by Using Decision-Theoretic Rough Set Model

There are many spectral bands or band functions developed for land-cover feature measurements. When the ratio of the number of training samples to the number of feature measurements is small, the traditional land-cover classifi cation is not accurate. To solve this problem, a decision-theoretic rough set model (DTRSM) is fi rst introduced. This model is linked with distinguishing different types of samples in the image. The samples in the minority classes will be misclassifi ed based on the model. To minimize the misclassifi cation, we propose an improved feature selection algorithm with comprehensive criteria. This algorithm is implemented on the Landsat TM data covering two disparate regions which are Lake Baiyangdian and Lake Qingpu located in the north and south of China, respectively. We compare the algorithm with other feature selection algorithms. Results show that the proposed method can effectively select key features for different data sets and the accuracy of classifi ers can be ensured.

An Improved Feature Selection Algorithm Based on Parzen Window and Conditional Mutual Information

In this paper, an improved feature selection algorithm by conditional mutual information with Parzen window was proposed, which adopted conditional mutual information as an evaluation criterion of feature selection in order to overcome the deficiency of feature redundant and used Parzen window to estimate the probability density functions and calculate the conditional mutual information of continuous variables, in such a way as to achieve feature selection for continuous data.

Spatial dependence matrix feature and redundancy elimination algorithm using AdaBoost for object detection

This paper describes a novel feature representation and selection approach for classification problems, especially for visual object detection within the framework of AdaBoost. This work is distinguished by two contributions. The first contribution is the introduction of a new feature generation and representation method called the spatial dependence matrix feature, which not only provides information related to the first-order statistics distribution of the object, but also gives some information about the relative positions within the object, more importantly, it can provide different degrees of importance for different discriminative parts within the object. It is flexible, extendable, and compatible with Haar-like features. The second contribution is an improved feature selection algorithm, which introduces a novel weighted features redundancy elimination rule that eliminates the irrelevant and redundant features from the candidate feature pool at every boosting stage when gradually training detector, and thus two advantages can be obtained: leading to selecting features with more discrimination and the final detector having a higher accuracy, and also increasing the learning convergence and achieving high training rates. Extensive experiments with synthetic and real scene data sets show that our works outperform conventional AdaBoost and are better than or at least equivalent to previously published results.