Feature Subset Selection Research Articles

Forecasting actual evapotranspiration without climate data based on stacked integration of DNN and meta-heuristic models across China from 1958 to 2021

As a non-linear phenomenon that varies along with agro-climatic conditions alongside many other factors, Evapotranspiration (ET) process represents a complexity when be assessed especially if there is a data scarcity in the weather data. However, even under such a data scarcity, the accurate estimates of ET values remain necessary for precise irrigation. So, the present study aims to: i) evaluate the performance of six hybrid machine learning (ML) models in estimating the monthly actual ET values under different agro-climatic conditions in China for seven provinces (Shandong, Jiangsu, Zhejiang, Fujian, Jiangxi, Hubei, and Henan), and ii) select the best-developed model based on statistical metrics and reduce errors between predicted and actual ET (AET) values. AET datasets were divided into 78% for model training (from 1958 to 2007) and the remaining was used for testing (from 2008 to 2021). Deep Neural Networks (DNN) was used as a standalone model at first then the stacking method was applied to integrate DNN with data-driven models such as Additive regression (AR), Random Forest (RF), Random Subspace (RSS), M5 Burned Tree (M5P) and Reduced Error Purning Tree (REPTree). Partial Auto-Correlation Function (PACF) was used for selection of the best lags inputs to the developed models. Results have revealed that DNN-based hybrid models held better performance than non-hybrid DNN models, such that the DNN-RF algorithm outperformed others during both training and testing stages, followed by DNN-RSS. This model has acquired the best values of every statistical measure [MAE (10.8, 12.9), RMSE (15.6, 17.4), RAE (31.9%, 41.4%), and RRSE (39.3%, 47.2%)] for training and testing, respectively. In contrast, the DNN model held the worst performance [MAE (14.9, 13.7), RMSE (20.1, 18.2), RAE (43.9%, 43.7%), and RRSE (50.6%, 49.3%)], for training and testing, respectively. Results from the study presented have revealed the capability of DNN-based hybrid models for long-term predictions of the AET values. Moreover, the DNN-RF model has been suggested as the most suitable model to improve future investigation for AET predictions, which could benefit the enhancement of the irrigation process and increase crop yield.

Sustainable Financial Fraud Detection Using Garra Rufa Fish Optimization Algorithm with Ensemble Deep Learning

Sustainable financial fraud detection (FD) comprises the use of sustainable and ethical practices in the detection of fraudulent activities in the financial sector. Credit card fraud (CCF) has dramatically increased with the advances in communication technology and e-commerce systems. Recently, deep learning (DL) and machine learning (ML) algorithms have been employed in CCF detection due to their features’ capability of building a powerful tool to find fraudulent transactions. With this motivation, this article focuses on designing an intelligent credit card fraud detection and classification system using the Garra Rufa Fish optimization algorithm with an ensemble-learning (CCFDC-GRFOEL) model. The CCFDC-GRFOEL model determines the presence of fraudulent and non-fraudulent credit card transactions via feature subset selection and an ensemble-learning process. To achieve this, the presented CCFDC-GRFOEL method derives a new GRFO-based feature subset selection (GRFO-FSS) approach for selecting a set of features. An ensemble-learning process, comprising an extreme learning machine (ELM), bidirectional long short-term memory (BiLSTM), and autoencoder (AE), is used for the detection of fraud transactions. Finally, the pelican optimization algorithm (POA) is used for parameter tuning of the three classifiers. The design of the GRFO-based feature selection and POA-based hyperparameter tuning of the ensemble models demonstrates the novelty of the work. The simulation results of the CCFDC-GRFOEL technique are tested on the credit card transaction dataset from the Kaggle repository and the results demonstrate the superiority of the CCFDC-GRFOEL technique over other existing approaches.

Comprehensive evaluation of surface parameter correlation in running-in wear process

The first wear process is running-in, which is a stage of each machine system must go through. The friction pair’s surface topography is an important feature of the running-in process. After running-in, the surface topography directly affects the performance in stable working state. The wear performance of the contact pair was focused on by mostly existing studies after a time; however, we should consider the correlation of the surface topography before and after running-in. In fact, exploration of this problem is significantly important to extend the service life of machine parts and improve the running performance. Based on the three-dimensional (3D) surface topography parameter evaluation method, in this study, experiments were designed to analyze the effect of the unworn surface topography on that after running-in, and investigated the correlation of 3D surface topography parameters. The results demonstrated that the topography characterization parameters, i.e., Sa, Sq, Sv, Sdc, Str, Sdq, and Sdr under different working conditions maintained a strong autocorrelation (the correlation coefficient of the same parameter). Moreover, the cross-correlation (the correlation coefficient between different parameters) among height parameters (Sq), hybrid parameters (Sdq, Sdr), and functional parameters (Sdc, Sk) was strong. The results of the surface roughness parameters found in this study can be used as input and output feature selection research based on working conditions and surface topography prediction after the modeling research of running-in surface topography.

Feature selection for image classification using ESSO

Feature selection is one of the major components of the data processing flow that correctly selects real-time entities for categorization. It is utilized in numerous research fields, including the machine learning approach. Feature selection and classification are beneficial to the processing of biomedical data in high-importance, high-dimensional datasets. Due to numerous obstacles in research, the current implementations are inadequate for predicting classification accuracy. To handle the classification challenge, we require dedicated neural network and classification model approaches. Initial features are selected using a method called "fuzzy c-means clustering with rough set theory" and are subsequently classified using "support vector machines." The learning machines' accuracy is seriously harmed by irrelevant and redundant features. The curse of dimensionality makes it difficult to locate data clusters in high-dimensional space. Data in the irrelevant dimensions may cause a lot of noise when the dimensionality grows. In addition, the current approach has a significant flaw: it is extremely time-consuming. The proposed solution used efficient feature subset selection in high-dimensional data to fix these problems. To begin with, our recommended method used Enhanced Social Spider Optimization (ESSO) computation, in which the standard Social Spider Optimization is improved with the aid of optimal radial based calculation to select the best highlights. When categorizing data, the Optimal Radial Basis Function Neural Network (ORBFNN) is used. Methods for calculating Artificial Bee Colony (ABC) are used to streamline RBFNN's sufficiency in characterizing smaller scale show information.

An evolutionary functional link artificial neural network for assessment of compressive strength of concrete structures

Compressive strength (CS) has been considered as the utmost critical parameter while designing concrete structures. Usually, it is determined through laboratory tests, which are expensive, time consuming, and requires consumptions of materials. Therefore, correct prediction of CS before actual placement of concrete is highly desirable. The relationship among the constituent materials that forms concrete structures is highly nonlinear and necessitates application of intelligent methods. Though a few such methods like artificial neural network (ANN) based models are available in the literature; their performance in the context is limited to certain extent and they have their own merits and demerits. Hence, to address some of the limitations of ANN (non-higher order Neural Network) based models, this contribution proposed a hybrid model in which a flat network i.e., a type of higher order neural network- functional link artificial neural network (FLANN) is used as the base structure and genetic algorithm (GA) is employed to find out the optimal FLANN parameters (i.e., GA + FLANN). The training process comprises selection of connection weights, bias, as well as optimal number of basis functions of FLANN by GA rather fixing them earlier. Thus, an optimal FLANN is crafted on fly from exploitation of training data. The proposed model is used to assess the CS of concrete cements from datasets available in the literature considering samples of curing age at 3, 7, 14, 28, 56, and 91 days. A rolling window is used for input selection and five evaluation metrics are used for performance evaluation. On an average, the GA + FLANN obtained 0.477052 MAPE (mean absolute percentage of error), 0.598067 ARV (average relative variance), 0.593862 UT (U of Theil’s statistic), 0.551452 NMSE (normalized mean squared error), and 0.206135 SD (standard deviation) values which are the lowest compared to others. The superiority of the model is established through comparative studies and statistical significance tests.

Improved Artificial Rabbits Optimization with Ensemble Learning-Based Traffic Flow Monitoring on Intelligent Transportation System

Traffic flow monitoring plays a crucial role in Intelligent Transportation Systems (ITS) by dealing with real-time data on traffic situations and allowing effectual traffic management and optimization. A typical approach used for traffic flow monitoring frequently depends on collection and analysis of the data through a manual process that is not only resource-intensive, but also a time-consuming process. Recently, Artificial Intelligence (AI) approaches like ensemble learning demonstrate promising outcomes in numerous ITS applications. With this stimulus, the current study proposes an Improved Artificial Rabbits Optimization with Ensemble Learning-based Traffic Flow Monitoring System (IAROEL-TFMS) for ITS. The primary intention of the proposed IAROEL-TFMS technique is to employ the feature subset selection process with optimal ensemble learning so as to predict the traffic flow. In order to accomplish this, the IAROEL-TFMS technique initially designs the IARO-based feature selection approach to elect a set of features. In addition, the traffic flow is predicted using the ensemble model that comprises a Gated Recurrent Unit (GRU), Long Short-term Memory (LSTM), and Bidirectional Gated Recurrent Unit (BiGRU). Finally, the Grasshopper Optimization Algorithm (GOA) is applied for the adjustment of the optimum hyperparameters of all three DL models. In order to highlight the improved prediction results of the proposed IAROEL-TFMS algorithm, an extensive range of simulations was conducted. The simulation outcomes imply the supremacy of the IAROEL-TFMS methodology over other existing approaches with a minimum RMSE of 16.4539.

Detection of Android Malware in the Internet of Things through the K-Nearest Neighbor Algorithm.

Predicting attacks in Android malware devices using machine learning for recommender systems-based IoT can be a challenging task. However, it is possible to use various machine-learning techniques to achieve this goal. An internet-based framework is used to predict and recommend Android malware on IoT devices. As the prevalence of Android devices grows, the malware creates new viruses on a regular basis, posing a threat to the central system's security and the privacy of the users. The suggested system uses static analysis to predict the malware in Android apps used by consumer devices. The training of the presented system is used to predict and recommend malicious devices to block them from transmitting the data to the cloud server. By taking into account various machine-learning methods, feature selection is performed and the K-Nearest Neighbor (KNN) machine-learning model is proposed. Testing was carried out on more than 10,000 Android applications to check malicious nodes and recommend that the cloud server block them. The developed model contemplated all four machine-learning algorithms in parallel, i.e., naive Bayes, decision tree, support vector machine, and the K-Nearest Neighbor approach and static analysis as a feature subset selection algorithm, and it achieved the highest prediction rate of 93% to predict the malware in real-world applications of consumer devices to minimize the utilization of energy. The experimental results show that KNN achieves 93%, 95%, 90%, and 92% accuracy, precision, recall and f1 measures, respectively.

Lung Cancer Detection by Multiple Feature Subset Extraction and Selection based on SVM-Weights and Genetic Algorithm-Neural Network

Lung Cancer Detection by Multiple Feature Subset Extraction and Selection based on SVM-Weights and Genetic Algorithm-Neural Network

Hybrid Hunter–Prey Optimization with Deep Learning-Based Fintech for Predicting Financial Crises in the Economy and Society

Financial technology (Fintech) plays a pivotal role in driving contemporary technology, society, economies, and many other fields. The new-generation Fintech is Smart Fintech, mainly empowered and inspired by data science and artificial intelligence (DSAI) technologies. Smart Fintech combines DSAI and transforms finance and economies for driving automated, intelligent, personalized financial and economic businesses, services and systems, and the whole of business. The strength and growth of the country’s economy were evaluated with the accurate prediction of how many companies will succeed and how many will fail. Financial crisis prediction (FCP) has a considerable effect on the economy. Prior research focuses mainly on deep learning (DL), machine learning (ML), and statistical approaches for forecasting the financial health of a company. Thus, this study presents a hybrid hunter–prey optimization with a deep learning-based FCP (HHPODL-FCP) technique. The objective of the HHPODL-FCP algorithm lies in the effective identification of the financial crisis in enterprises or organizations. To accomplish this, the HHPODL-FCP method makes use of the HHPO algorithm for the feature subset selection process. In addition, the HHPODL-FCP technique employs the gated attention recurrent network (GARN) model for the identification and classification of financial and non-financial crises. The HHPODL-FCP method exploits a sparrow search algorithm (SSA)-based hyperparameter tuning process to enrich the performance of the GARN model. The simulation results of the HHPODL-FCP method are tested on different financial datasets. A wide range of experiments highlighted the remarkable performance of the HHPODL-FCP method over recent techniques under various measures.

Linear adversarial vector modeling based intrusion detection using feature subset selection and representation methods

Linear adversarial vector modeling based intrusion detection using feature subset selection and representation methods

A two-stage feature selection approach using hybrid quasi-opposition self-adaptive coati optimization algorithm for breast cancer classification

Breast cancer (BC) is one of the leading causes of high mortality rates among women. An early disease diagnosis is crucial in breast cancer’s treatment for improving the survival rate of people who have been afflicted. Human error, incorrect diagnoses, and insufficient time make manual techniques for breast cancer diagnosis ineffective and inefficient. An automated breast cancer diagnosis system assists the medical expert in diagnosing the cancer at its earliest stage. When attempting to investigate all aspects (high dimensionality) of the medical records, the automated diagnosis system faces significant challenges. It reduces the success rate of the diagnosis system. This study presents a two-stage feature subset selection approach using quasi-opposition self-adaptive coati optimization algorithm for breast cancer classification. The proposed scheme’s fitness function aims to maximize the success rate and minimize the feature selection ratio. We used the three benchmark datasets (Breast Cancer Wisconsin Diagnostic Dataset (WDBC), Breast Cancer Wisconsin Prognostic Dataset (WPBC), and Breast Cancer Wisconsin Original Dataset (WBCD)) to examine the proposed scheme’s performance. The comparison analysis results show that the proposed scheme outperforms the other competitor schemes.

Feature subset selection in data-stream environments using asymmetric hidden Markov models and novelty detection

With the increase of computational power and memory capacity, it is possible to record and analyse lots of features of different nature in real time or in a data stream manner. Nonetheless, in many applications, not all variables may be relevant to explain the nature of the data. Feature subset selection strategies are therefore required to extract the relevant features to understand the data generating process. Much work has been performed in the area of supervised classification problems regarding feature selection, whereas for unsupervised environments, scarce studies have been conducted. In current feature subset selection methodologies for unsupervised data streams, the set of relevant variables are reevaluated or forgotten whenever a new instance or chunk of data arrives: this approach can be computationally expensive or unnecessary. Therefore, in this article, we provide a method to perform feature subset selection in unsupervised data streams. An embedded feature subset selection methodology based on asymmetric hidden Markov models and novel concept discovery strategies is used. This methodology can be used to detect novel concepts in data, and update the set of relevant features when a drift in the data stream is detected. Thus, the relevant variables are updated only when needed. To make this possible, we provide a model for batch and online problems, that is capable of dynamically determining the relevant variables to address feature selection in data streams. Additionally, the model provides domain insights using context-specific Bayesian networks which can be helpful to understand the dynamic process. To validate the proposed methodology, synthetic and real data from ball-bearings are used. Additionally, the proposed methodology is compared with other state of the art methodologies. The proposed method can be used to update the relevant features when needed and is more stable than its competitors.

Leveraging Machine Learning for Pipeline Condition Assessment

Pipeline condition assessment is a cost-effective method to determine the status of pipeline structure and predict failure probability. Although 100% inspection may not be feasible for decision makers, recent advancements in machine learning techniques have enabled more effective pipeline condition assessment. This paper provides a comprehensive review of machine learning applications in pipeline condition assessment, covering aspects such as fault diagnosis, risk prediction, parameter prediction, and visual defect recognition. The present study endeavors to make the following contributions: (1) extraction of the model, data size, and other relevant information from 91 papers, (2) in-depth analysis of the state of the art and frameworks of the models discussed in the 91 papers, (3) summary of the data characteristics, input variables, and accuracy of machine learning models, and (4) exploration of the potential avenues for future research in the use of machine learning for pipeline condition assessment. This review aims to serve as a practical reference for scholars engaged in related research. The review highlights the fact that the majority of the models employed in pipeline condition assessment are original, and the utilization of hybrid models remains limited. Transfer learning and reinforcement learning are identified as potential avenues for future research because they hold promise in facilitating the adaptive selection of model inputs and the transfer of models to similar projects. Furthermore, breaking down data barriers is deemed essential for advancing the use of machine learning in pipeline condition assessment.

The Eighteenth Data Release of the Sloan Digital Sky Surveys: Targeting and First Spectra from SDSS-V

The eighteenth data release (DR18) of the Sloan Digital Sky Survey (SDSS) is the first one for SDSS-V, the fifth generation of the survey. SDSS-V comprises three primary scientific programs or “Mappers”: the Milky Way Mapper (MWM), the Black Hole Mapper (BHM), and the Local Volume Mapper. This data release contains extensive targeting information for the two multiobject spectroscopy programs (MWM and BHM), including input catalogs and selection functions for their numerous scientific objectives. We describe the production of the targeting databases and their calibration and scientifically focused components. DR18 also includes ∼25,000 new SDSS spectra and supplemental information for X-ray sources identified by eROSITA in its eFEDS field. We present updates to some of the SDSS software pipelines and preview changes anticipated for DR19. We also describe three value-added catalogs (VACs) based on SDSS-IV data that have been published since DR17, and one VAC based on the SDSS-V data in the eFEDS field.

AutoFe-Sel: A Meta-learning based methodology for Recommending Feature Subset Selection Algorithms

AutoFe-Sel: A Meta-learning based methodology for Recommending Feature Subset Selection Algorithms

Improved Gaussian mixture model to map the flooded crops of VV and VH polarization data

Accurate and timely monitoring of flooded crop areas is crucial for disaster rescue and loss assessment. However, most flooded crop monitoring methods based on synthetic aperture radar (SAR) imagery were developed for rice, which is probably inappropriate for crops with complex canopy structures that strongly attenuate SAR signals. Additionally, these methods often rely on empirical thresholds and region-specific reference samples, limiting their reliability and applicability on a larger spatial scale. To address these issues, we developed a novel flooded crop mapping approach at a regional scale using Sentinel-1 time-series data and an unsupervised Gaussian Mixture Model (GMM). Our approach leverages a Flood Separability Index (FSI) derived from the fitted probability density function of flooded and non-flooded crop areas in a GMM. This allows us to overcome the limitations of manual input selection in previous studies. The multi-temporal GMM was constructed using the time-series images with optimal polarization to estimate the flooded crop extents on a regional scale. We also investigated the scattering mechanisms of three typical crop disaster structures within an agricultural landscape area. Our results indicate that the proposed multi-temporal GMM is robust in crop planting areas with complex canopy structures. The performance of both single-temporal and multi-temporal GMMs surpasses that of baseline methods such as Otsu and K-means. Compared with VV polarization, VH polarization exhibits greater potential for accurately mapping flooded crops in complex agricultural regions. Our approach does not require labeled samples or many predefined parameters, making it fast and feasible for mapping flooded crops with complex canopy structures in large spatial areas.

Feature subset selection for data and feature streams: a review

Real-world problems are commonly characterized by a high feature dimensionality, which hinders the modelling and descriptive analysis of the data. However, some of these data may be irrelevant or redundant for the learning process. Different approaches can be used to reduce this information, improving not only the speed of building models but also their performance and interpretability. In this review, we focus on feature subset selection (FSS) techniques, which select a subset of the original feature set without making any transformation on the attributes. Traditional batch FSS algorithms may not be adequate to efficiently handle large volumes of data, either because memory problems arise or data are received in a sequential manner. Thus, this article aims to survey the state of the art of incremental FSS algorithms, which can perform more efficiently under these circumstances. Different strategies are described, such as incrementally updating feature weights, applying information theory or using rough set-based FSS, as well as multiple supervised and unsupervised learning tasks where the application of FSS is interesting.

Directional modifier adaptation based on input selection for real-time optimization

Modifier adaptation (MA) is commonly used for economic optimization of systems under model uncertainty. In MA, gradient correction terms require estimation through perturbations, thus delaying the optimization procedure. A directional modifier adaptation method is proposed whereby a subset of available gradient corrections are made. An algorithm that evaluates possible adaptation strategies and chooses those with the largest economic effect is proposed, thereby allowing economical operation with less delay. The proposed scheme, named dMAIS, is deployed on the Williams-Otto process where it is found to outperform MA if not inhibited by filtering. Systems can also suffer from constraint violation if uncertainty is present, hampering safety and profitability. An adjustment step is proposed as part of dMAIS, whereby gradients are used to drive the plant to constraint satisfaction. The adjustments are studied in an evaporator case with a product quality constraint whereby dMAIS is shown to result in infrequent violations leading to higher throughput. The proposed approach was also compared to standard directional MA in the evaporator case study, where it was found to be economically beneficial. The benefits of dMAIS are observed most saliently for systems with increasing disturbance frequency.

Comparative Analysis of Data Mining Techniques for Heart Disease Prediction: A Focus on Neural Networks and Decision Trees

Heart disease is a general term used to describe numerous medical conditions that directly affect the heart and its various components. It is a prevalent health concern in modern times. The focus of this paper is to evaluate different data mining techniques for the prediction of heart disease, which have been introduced in recent years. The findings indicate that neural networks using 15 attributes demonstrate the best performance among all other data mining techniques. Additionally, the analysis concludes that decision trees, with the assistance of genetic algorithms and feature subset selection, also exhibit high accuracy. The study concludes that data mining techniques can effectively predict heart disease and that the choice of technique depends on the specific context of the analysis. The study suggests that decision trees and artificial neural network models are suitable for heart disease prediction. The study also recommends further research to explore the use of other data mining techniques for heart disease prediction.

Filter-PCA-Based Process Monitoring and Defect Identification During Climbing Helium Arc Welding Process Using DE-SVM

In-process defect detection for climbing helium arc welding of aluminum alloy is challenging due to the highly complex nature of the heat and mass transfer process. This paper presents a process monitoring and defects identification method for weld quality monitoring during climbing helium arc welding based on molten pool visual sensing. High contrast molten pool images were acquired by the designed passive vision sensor. Then, an image processing algorithm based on Otsu's method and visual saliency features was proposed to extract multi-region features from the acquired images in real-time. An optimal feature subset selection approach based on filter and principal component analysis(Filter-PCA) was developed to select the components related to welding defects. Finally, a novel prediction model using Filter-PCA and the support vector machine based on differential evolution algorithm(DE-SVM) was established to identify undercut weld, snake-like weld, and sound weld. The experiment results obtained in the actual production line of 2219 aluminum alloy rocket propellent tank show that the proposed method is remarkably effective and robust. This research provides a technical approach for early identification and early warning of defects during climbing helium arc welding of aluminum alloy. It also provides a basis for stability and online control of weld formation quality under time-varying welding pose.