Search

When contemplating the improvement of overall performance in machine learning (ML) models, a critical strategy for optimizing data preparation is feature selection (FS). There has been a significant rise in the popularity of metaheuristic FS algorithms in recent times. This can be attributed to their proficiency in accurately identifying and selecting the most relevant features for ML tasks. This study presents three feature selection strategies that utilize metaheuristic algorithms. The methodologies mentioned include the Gravitational Search Optimization Algorithm (GSA), Emperor Penguin Optimization (EPO), and a hybrid approach of GSA and EPO referred to as hGSAEPO. Previous research has explored the use of baseline algorithms for feature selection in various ML tasks. However, there is a lack of investigation regarding their application specifically in breast cancer(BC) classification. A combination of these two has been utilized for the first occasion. The purpose of selecting BC as the study of investigation is due to the reason that this illness is recognized as the second most prevalent cause of mortality in the female population. If the condition is detected in its initial phases, it can be remedied and can assist individuals in evading superfluous medical processes. The procedure of selecting relevant features holds significant importance in the purpose of predicting ailments like BC. The current research presents an innovative methodology that employs three soft-computing algorithms, EPO, GSA, and their proposed hybrid hGSAEPO to efficiently identify significant features while concurrently decreasing the occurrence of irrelevant ones, simplifying overall complexity and enhancing the accuracy. The utilization of these soft computing methodologies and six ML classifiers presents a viable framework for prognostic research through the classification of data instances on Wisconsin Diagnostic Breast Cancer (WDBC). The experimental findings of eight experiments conducted suggest that the suggested approach exhibits exceptional performance in the context of binary classification for BC by computing astounding results like precision of 0.9800, sensitivity of 0.9700, specificity of 0.9887, F1-score of 0.9539, area under the curve(AUC) surpassing 0.998, with an accuracy of 98.31%. We achieved our objectives by presenting a dependable clinical prediction system for healthcare professionals for efficient diagnosis. [ABSTRACT FROM AUTHOR]

A brain-computer interface (BCI) based on an electroencephalograph (EEG) establishes a new channel of communication between the human brain and a computer. Redundant, noisy, and irrelevant channels lead to high computational costs and poor classification accuracy. Therefore, an effective feature selection technique for determining the optimal number of channels can improve BCI's performance. However, existing meta-heuristic algorithms are prone to get trapped in local optimum due to high dimensional dataset. Thus, to reduce dimension, solve inter subject variation and choose an optimal subset of channels, a novel framework called Component Loading followed by Clustering and Classification (CLCC) is proposed in this paper. This novel framework is further divided into two experiment configurations-CLCC with Feature Selection (CLCC-FS) and CLCC without Feature Selection (CLCC-WFS). All these frameworks have been implemented on a motor imagery (MI) EEG dataset of 10 subjects in order to choose the best subset of channels. Further, seven different classifiers have been employed to assess the performance. Experimental outcomes show that on comparing various feature selection techniques, our proposed algorithm i.e., CLCC-FS Opposition-Based Whale Optimization Algorithm (CLCC-FS(OBWOA)) performed substantially better than the other feature selection techniques. We demonstrate that the proposed algorithm is able to achieve 99.6% accuracy by using only few channels and can improve the practicality of the BCI system by reducing the computation cost. [ABSTRACT FROM AUTHOR]

Feature selection plays a critical role in reducing high-dimensional feature space in machine learning applications without affecting the accuracy of performance. The feature selection problem has been extensively studied in the literature. Nevertheless, few studies have been conducted on unsupervised text feature selection because of the absence of feature class labels and local optimization limitations. For that, this paper proposes three binary versions of the fire hawk optimizer based on different transfer functions for unsupervised text feature selection that can be used to select the most informative features for text clustering. The internal feature subset was evaluated using a mean absolute difference filter. The performance of the proposed methods is tested and compared with other state-of-the-art metaheuristic algorithms on several different benchmark text datasets from different sources using various evaluation metrics. The K-means clustering is applied to cluster documents based on the features selected by the methods. The results of the experiments indicate the effectiveness of the proposed method based on the S-shaped transfer function for improving the performance of document clustering compared with other feature selection methods. [ABSTRACT FROM AUTHOR]

This study investigates the application of deep learning, ensemble learning, metaheuristic optimization, and image processing techniques for detecting lung and colon cancers, aiming to enhance treatment efficacy and improve survival rates. We introduce a metaheuristic-driven two-stage ensemble deep learning model for efficient lung/colon cancer classification. The diagnosis of lung and colon cancers is attempted using several unique indicators by different versions of deep Convolutional Neural Networks (CNNs) in feature extraction and model constructions, and utilizing the power of various Machine Learning (ML) algorithms for final classification. Specifically, we consider different scenarios consisting of two-class colon cancer, three-class lung cancer, and five-class combined lung/colon cancer to conduct feature extraction using four CNNs. These extracted features are then integrated to create a comprehensive feature set. In the next step, the optimization of the feature selection is conducted using a metaheuristic algorithm based on the Electric Eel Foraging Optimization (EEFO). This optimized feature subset is subsequently employed in various ML algorithms to determine the most effective ones through a rigorous evaluation process. The top-performing algorithms are refined using the High-Performance Filter (HPF) and integrated into an ensemble learning framework employing weighted averaging. Our findings indicate that the proposed ensemble learning model significantly surpasses existing methods in classification accuracy across all datasets, achieving accuracies of 99.85% for the two-class, 98.70% for the three-class, and 98.96% for the five-class datasets. [ABSTRACT FROM AUTHOR]

Cardio Vascular Diseases (CVD) has become a serious concern for humans as fatalities rate due to CVD are increasing at an alarming pace. With the aid of machine learning techniques, heart illnesses can be predicted much earlier, and therapy or dietary changes can prevent deaths. By combining predictions from various individual models, the machine learning technique known as ensemble learning improves forecasting accuracy and resiliency. In this work, a Super Learner Ensemble Model is used where the base learners are a diverse combination of linear, probabilistic, bagging, boosting and stacking models. To improve the performance of the Super Learner Ensemble Model, an Optimized Super Learner Ensemble Model (OSLEM) is proposed, where optimal selection of base learners in the ensemble is done based on the pairwise disagreement accuracy diversity measure of classifiers in each best fitness whale obtained by different iterations of Whale Optimization Algorithm (WOA). ModifiedBoostARoota (MBAR), a wrapper feature selection technique is used to choose the most significant features of six different heart datasets and the proposed OSLEM modelled on the selected features exhibits high performance compared to other existing ensemble models. [ABSTRACT FROM AUTHOR]

The early detection of cancer is of paramount importance in the medical field, as it can lead to more precise and effective interventions for successful cancer treatments. Cancer datasets typically contain numerous gene expression levels as features but with a limited number of samples. Thus, feature selection is a crucial initial step to streamline prediction algorithms. These selected features, or genes, play a pivotal role in cancer identification, treatment selection, and variation analysis among different techniques. To address this challenge, present two novel methodologies, by combining Cuckoo Search (CS) and Spider Monkey Optimization (SMO), referred to as SMOCS (Cuckoo Search followed by Spider Monkey Optimization) and CSSMO (Spider Monkey Optimization followed by Cuckoo Search). These approaches are designed for harnessing the strengths of both metaheuristic algorithms to identify a subset of genes that aid in early-stage cancer prediction. Additionally, to enhance the accuracy of the both algorithms, we employ a gene expression reduction method known as minimum Redundancy Maximum Relevance (mRMR) to reduce redundancy in cancer datasets. Subsequently, these gene subsets are classified using Deep Learning (DL) to identify distinct groups or classes associated with specific cancer types. We evaluate the performance of our proposed approaches using six different cancer datasets, assessing cancer sample classification and prediction through metrics such as Recall, Precision, F1-Score, and confusion matrix analysis. Our gene selection methods, in conjunction with DL, achieves significantly improved prediction accuracy when applied to large gene expression datasets compared to existing Deep Learning (DL) and Machine learning models. Experimental results shows that both SMOCS and CSSMO tend to classify cancer with high prediction accuracy, but SMOCS algorithm gives higher prediction accuracy for all the six datasets used with a maximum accuracy of 100%. [ABSTRACT FROM AUTHOR]

Breast cancer is currently one of the most prevalent cancers affecting women globally. Uncontrolled growth and division of breast cells lead to the formation of tumors, marking the onset of breast cancer. Predicting breast cancer is essential for early detection, making treatment plans, and implementing preventive measures, ultimately improving patient outcomes and reducing mortality rates. In recent years, numerous studies have been published to predict breast cancer where researchers use a variety of methods. Most investigations have been conducted using narrow and specific datasets, often resulting in a lack of accuracy. Such methods may not be suitable for clinical use. The study aims to address the limitations of existing models in terms of robustness and generalization across diverse datasets. In our study, we employed two metaheuristic algorithms, namely, genetic algorithm (GA) and chemical reaction optimization (CRO) with machine learning techniques, including support vector machine (SVM), decision tree, random forest, and XGBoost. GA and CRO are used to optimize the feature selection process. It enables machine learning algorithms to predict more accurately. Experiments were conducted on three datasets, namely, Wisconsin Breast Cancer (WBC), Breast Cancer‐the University of California, Irvine (BC‐UCI), and Breast Cancer Coimbra (BCC) datasets. The datasets contain 569, 286, and 116 instances, respectively. The classifiers with optimized features consistently outperformed the classifiers without feature optimization in terms of accuracy, precision, recall, specificity, and F1 score. Among the compared methods published recently, our method attained the highest accuracies of 99.64% in the WBC dataset and 98% in the BCC dataset, as well as the second highest accuracy of 99.12% in the BC‐UCI dataset. Comparative analysis demonstrated the superiority of our approach over existing methods. [ABSTRACT FROM AUTHOR]

Feature selection (FS) is a significant dimensionality reduction technique in machine learning and data mining that is adept at managing high-dimensional data efficiently and enhancing model performance. Metaheuristic algorithms have become one of the most promising solutions in FS owing to their powerful search capabilities as well as their performance. In this paper, the novel improved binary walrus optimizer (WO) algorithm utilizing the golden sine strategy, elite opposition-based learning (EOBL), and population regeneration mechanism (BGEPWO) is proposed for FS. First, the population is initialized using an iterative chaotic map with infinite collapses (ICMIC) chaotic map to improve the diversity. Second, a safe signal is obtained by introducing an adaptive operator to enhance the stability of the WO and optimize the trade-off between exploration and exploitation of the algorithm. Third, BGEPWO innovatively designs a population regeneration mechanism to continuously eliminate hopeless individuals and generate new promising ones, which keeps the population moving toward the optimal solution and accelerates the convergence process. Fourth, EOBL is used to guide the escape behavior of the walrus to expand the search range. Finally, the golden sine strategy is utilized for perturbing the population in the late iteration to improve the algorithm's capacity to evade local optima. The BGEPWO algorithm underwent evaluation on 21 datasets of different sizes and was compared with the BWO algorithm and 10 other representative optimization algorithms. The experimental results demonstrate that BGEPWO outperforms these competing algorithms in terms of fitness value, number of selected features, and F1-score in most datasets. The proposed algorithm achieves higher accuracy, better feature reduction ability, and stronger convergence by increasing population diversity, continuously balancing exploration and exploitation processes and effectively escaping local optimal traps. [ABSTRACT FROM AUTHOR]

Feature selection techniques play a vital role in the processes that deal with enormous amounts of data. These techniques have become extremely crucial and necessary for data mining and machine learning problems. Researchers have always been in a race to develop and provide libraries and frameworks to standardise this procedure. In this work, we propose a hybrid meta‐heuristic algorithm to facilitate the problem of feature selection for classification problems in machine learning. It is a python based, lucid and efficient algorithm geared towards optimising and striking a balance between the number of features selected and accuracy. The proposed work is a binary hybrid of existing meta‐heuristic algorithms, the particle swarm optimisation (PSO) algorithm, and the firefly algorithm (FA) such that it blends the best of each algorithm to provide an optimised and efficient way of solving the said problem. The suggested approach is assessed against six datasets from different domains that are publicly available at the UCI repository to demonstrate its validity. The datasets are Breast cancer, Iris, WBC, Mushroom, Glass ID, and Abalone. This approach has also been evaluated against similar, such evolutionary‐based approaches to prove its superiority. Various metrics such as accuracy, precision, recall, f1 score, number of selected features, and run time have been analysed, measured, and compared. The hybrid firefly particle swarm optimisation algorithm is found to be suitable for feature selection problems. [ABSTRACT FROM AUTHOR]

Optimization techniques, particularly meta-heuristic algorithms, are highly effective in optimizing and enhancing efficiency across diverse models and systems, renowned for their ability to attain optimal or near-optimal solutions within a reasonable timeframe. In this work, the Puma Optimizer (PO) is proposed as a new optimization algorithm inspired from the intelligence and life of Pumas in. In this algorithm, unique and powerful mechanisms have been proposed in each phase of exploration and exploitation, which has increased the algorithm's performance against all kinds of optimization problems. In addition, a new type of intelligent mechanism, which is a type of hyper-heuristic for phase change, is presented. Using this mechanism, the PO algorithm can perform a phase change operation during the optimization operation and balance both phases. Each phase is automatically adjusted to the nature of the problem. To evaluate the proposed algorithm, 23 standard functions and CEC2019 functions were used and compared with different types of optimization algorithms. Moreover, using the statistical test T-test and the execution time to solve the problem have been discussed. Finally, it has been tested using four machine learning and data mining problems, and the results obtained from all the analysis signifies the excellent performance of this algorithm against all kinds of problems compared to other optimizers. This algorithm has performed better than the compared algorithms in 27 benchmarks out of 33 benchmarks and has obtained better results in solving the clustering problem in 7 data sets out of 10 data sets. Furthermore, the results obtained in the problems of community detection and feature selection and MLP were superior. The source codes of the PO algorithm are publicly available at https://www.mathworks.com/matlabcentral/fileexchange/157231-puma-optimizer-po. [ABSTRACT FROM AUTHOR]

Learning from imbalanced data is a challenging task in the machine learning field, as with this type of data, many traditional supervised learning algorithms tend to focus more on the majority class while damaging the interests of the minority class. Stacking ensemble, which formulates an ensemble by using a meta-learner to combine the predictions of multiple base classifiers, has been used for solving class imbalance learning issues. Specifically, in the context of class imbalance learning, a stacking ensemble learning algorithm is generally considered to combine with a specific sampling algorithm. Such an operation, however, might suffer from suboptimization problems as only using a sampling strategy may make it difficult to acquire diverse enough features. In addition, we also note that using all of these features may damage the meta-learner as there may exist noisy and redundant features. To address these problems, we have proposed a novel stacking ensemble learning algorithm named MSFSS, which divides the learning procedure into two phases. The first stage combined multiple sampling algorithms and multiple supervised learning approaches to construct meta feature space by means of cross combination. The adoption of this strategy satisfied the diversity of the stacking ensemble. The second phase adopted the whale optimization algorithm (WOA) to select the optimal sub-feature combination from the meta feature space, which further improved the quality of the features. Finally, a linear regression classifier was trained as the meta learner to conduct the final prediction. Experimental results on 40 benchmarked imbalanced datasets showed that the proposed MSFSS algorithm significantly outperformed several popular and state-of-the-art class imbalance ensemble learning algorithms. Specifically, the MSFSS acquired the best results in terms of the F-measure metric on 27 datasets and the best results in terms of the G-mean metric on 26 datasets, out of 40 datasets. Although it required consuming more time than several other competitors, the increment of the running time was acceptable. The experimental results indicated the effectiveness and superiority of the proposed MSFSS algorithm. [ABSTRACT FROM AUTHOR]

Cardiovascular diseases (CVD) are a leading cause of mortality globally, necessitating the development of efficient diagnostic tools. Machine learning (ML) and metaheuristic algorithms have become prevalent in addressing these challenges, providing promising solutions in medical diagnostics. However, traditional ML approaches often need to be improved in feature selection and optimization, leading to suboptimal performance in complex diagnostic tasks. To overcome these limitations, this study introduces a new hybrid method called CSA-DE-LR, which combines the clonal selection algorithm (CSA) and differential evolution (DE) with logistic regression. This integration is designed to optimize logistic regression weights efficiently for the accurate classification of CVD. The methodology employs three optimization strategies based on the F1 score, the Matthews correlation coefficient (MCC), and the mean absolute error (MAE). Extensive evaluations on benchmark datasets, namely Cleveland and Statlog, reveal that CSA-DE-LR outperforms state-of-the-art ML methods. In addition, generalization is evaluated using the Breast Cancer Wisconsin Original (WBCO) and Breast Cancer Wisconsin Diagnostic (WBCD) datasets. Significantly, the proposed model demonstrates superior efficacy compared to previous research studies in this domain. This study's findings highlight the potential of hybrid machine learning approaches for improving diagnostic accuracy, offering a significant advancement in the fields of medical data analysis and CVD diagnosis. [ABSTRACT FROM AUTHOR]

The dropout rate in underdeveloped and emerging countries is a pressing social issue, as highlighted by studies conducted by The Organization for Economic Co-operation and Development. This study compares five feature selection techniques to address this challenge and improve the automatic detection of dropout risk. The methodological design involves three distinct phases: data preparation, feature selection, and model evaluation utilizing machine learning algorithms. The results demonstrate that (1) the top features identified by feature selection techniques, i.e., those constructed through feature engineering, proved to be among the most effective in classifying student dropout; (2) the F-score of the best model increased by 5% with feature selection techniques; and (3) depending on the type of feature selection, the performance of the machine learning algorithm can vary, potentially increasing or decreasing based on the sensitivity of features with higher noise. At the same time, metaheuristic algorithms demonstrated significant precision improvements, but there was a risk of increasing errors and reducing recall. [ABSTRACT FROM AUTHOR]

Intrusion detection systems (IDSs) in wireless sensor networks (WSNs) rely heavily on effective feature selection (FS) for enhanced efficacy. This study proposes a novel approach called Genetic Sacrificial Whale Optimization (GSWO) to address the limitations of conventional methods. GSWO combines a genetic algorithm (GA) and whale optimization algorithms (WOA) modified by applying a new three-population division strategy with a proposed conditional inherited choice (CIC) to overcome premature convergence in WOA. The proposed approach achieves a balance between exploration and exploitation and enhances global search abilities. Additionally, the CatBoost model is employed for classification, effectively handling categorical data with complex patterns. A new technique for fine-tuning CatBoost's hyperparameters is introduced, using effective quantization and the GSWO strategy. Extensive experimentation on various datasets demonstrates the superiority of GSWO-CatBoost, achieving higher accuracy rates on the WSN-DS, WSNBFSF, NSL-KDD, and CICIDS2017 datasets than the existing approaches. The comprehensive evaluations highlight the real-time applicability and accuracy of the proposed method across diverse data sources, including specialized WSN datasets and established benchmarks. Specifically, our GSWO-CatBoost method has an inference time nearly 100 times faster than deep learning methods while achieving high accuracy rates of 99.65%, 99.99%, 99.76%, and 99.74% for WSN-DS, WSNBFSF, NSL-KDD, and CICIDS2017, respectively. [ABSTRACT FROM AUTHOR]

Data categorization is a top concern in medical data to predict and detect illnesses; thus, it is applied in modern healthcare informatics. In modern informatics, machine learning and deep learning models have enjoyed great attention for categorizing medical data and improving illness detection. However, the existing techniques, such as features with high dimensionality, computational complexity, and long-term execution duration, raise fundamental problems. This study presents a novel classification model employing metaheuristic methods to maximize efficient positives on Chronic Kidney Disease diagnosis. The medical data is initially massively pre-processed, where the data is purified with various mechanisms, including missing values resolution, data transformation, and the employment of normalization procedures. The focus of such processes is to leverage the handling of the missing values and prepare the data for deep analysis. We adopt the Binary Grey Wolf Optimization method, a reliable subset selection feature using metaheuristics. This operation is aimed at improving illness prediction accuracy. In the classification step, the model adopts the Extreme Learning Machine with hidden nodes through data optimization to predict the presence of CKD. The complete classifier evaluation employs established measures, including recall, specificity, kappa, F-score, and accuracy, in addition to the feature selection. Data related to the study show that the proposed approach records high levels of accuracy, which is better than the existing models. [ABSTRACT FROM AUTHOR]

Summary: Machine learning approaches are widely used for the detection and classification of emerging botnet variations due to their ability to yield more precise results compared to traditional methods. The relevancy of the features plays a major role in these detection algorithms' effectiveness. As such, the most distinctive characteristics must be extracted from a high‐dimensional dataset that is used to classify botnets. Nevertheless, we discovered that the majority of earlier studies lacked proper analysis and paid little attention to the various feature selection techniques. The main goal of this work is to investigate and assess the advantages and disadvantages of the different feature selection techniques used for botnet detection. Studies show that feature selection is a very efficient way to decrease the amount of storage and processing power required while simultaneously increasing classification accuracy. As a consequence, its application in many other fields has grown. The field of feature selection is recognized for its non‐deterministic polynomial‐time hardness; to mitigate this hardness, metaheuristic techniques have been applied. Metaheuristic algorithms are exceptionally good at performing a global search. In order to choose feature subsets optimally in the field of botnet detection, we additionally prioritize the use of metaheuristic methods. This study offers a more thorough insight of the feature selection strategies that are primarily employed by machine learning‐based botnet detection models. It also offers insights into how better feature selection approaches might be applied to strengthen botnet detection mechanisms. Additionally, it will help in understanding the limitations of existing approaches and identifying areas for improvement. [ABSTRACT FROM AUTHOR]

Machine learning classifiers have been widely used for crop mapping; however, employing numerous features can reduce classifier accuracy. Feature selection is a widely recognized technique that addresses the issue of high dimensionality by choosing the most relevant subset of features with the highest degree of relevance and practicality. Therefore, the optimal combination of feature selection methods and classifiers for high-precision crop mapping remains an open problem. In this study, we introduce a novel hybrid feature selection approach to obtain the optimal subset of features for effective crop mapping. Initially, spectro-temporal remote sensing features from the dataset were ranked using two distinct feature selection techniques: Mutual Information (MI) and ReliefF filter. The most relevant features from each filter-based approach were combined into a union subset. Subsequently, the Gray Wolf Optimization (GWO), a metaheuristic optimization technique, was applied to enhance the feature set generated in the initial step. In the final stage, a random forest classifier leveraged the optimized feature subset for accurate crop type prediction. The effectiveness of the proposed approach was evaluated in Behiera governorate, Egypt. Comparative analysis against existing crop mapping methods demonstrated the superior performance of the proposed approach, achieving an accuracy rate of 82%. Furthermore, this approach showcased a significant reduction in the feature count, streamlining the set from 308 down to only 26 features. [ABSTRACT FROM AUTHOR]

Cardiovascular disease (CVD) detection using deep learning (DL) includes leveraging advanced neural network (NN) models to analyze medical data, namely imaging, electrocardiograms (ECGs), and patient records. This study introduces a new Nature Inspired Metaheuristic Algorithm with Deep Learning for Healthcare Data Analysis (NIMADL-HDA) technique. The NIMADL-HDA technique examines healthcare data for the recognition and classification of CVD. In the presented NIMADL-HDA technique, Z-score normalization was initially performed to normalize the input data. In addition, the NIMADL-HDA method made use of a barnacle mating optimizer (BMO) for the feature selection (FS) process. For healthcare data classification, a convolutional long short-term memory (CLSTM) model was employed. At last, the prairie dog optimization (PDO) algorithm was exploited for the optimal hyperparameter selection procedure. The experimentation outcome analysis of the NIMADL-HDA technique was verified on a benchmark healthcare dataset. The obtained outcomes stated that the NIMADL-HDA technique reached an effectual performance over other models. The NIMADL-HDA method provides an adaptable and sophisticated solution for healthcare data analysis, aiming to improve the interpretability and accuracy of the algorithm in terms of medical applications. [ABSTRACT FROM AUTHOR]

The threat of malicious software has evolved into a major concern regarding the security of the system and network infrastructure. Machine learning algorithms have been successfully utilized to classify malware files into malicious or benign. However, the exponential growth in data volume and feature dimensionality poses challenges for machine learning, resulting in reduced classification accuracy and heightened computational costs. Feature selection is an essential process that can address these challenges by eliminating irrelevant, redundant, and less informative features that may adversely affect classifier performance. In this study, we introduce an enhanced Whale Optimization Algorithm (EWOA) aimed at improving classification accuracy, feature selection, and overall malware detection model efficiency. The proposed EWOA introduces an enhanced search mechanism that integrates mutation and neighborhood search strategies, aiming to refine its exploration strategy. This novel approach is more adept at steering clear of local optima. Additionally, EWOA augments its population diversity by incorporating the Opposite-Based Learning technique (OBL) during its initial phase. To assess the efficacy of the proposed method, performance evaluations were conducted using the CIC-MalMem-2022 dataset. Various aspects including the number of features, efficiency, fitness value, accuracy, and statistical tests were compared across different optimization algorithms: Gray Wolf Optimization Algorithm (GOA), Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Artificial Lion Optimization (ALO), Butterfly Optimization Algorithm (BOA), and Slime Mould Algorithm (SMA). The experimental results affirm the superiority of EWOA over other optimization algorithms in diverse areas, such as classification accuracy (99.987%), fitness value (0.00084511%), and average feature count (on average, 3.97 features). [ABSTRACT FROM AUTHOR]

Bathymetry data is indispensable for a variety of aquatic field studies and benthic resource inventories. Determining water depth can be accomplished through an echo sounding system or remote estimation utilizing space-borne and air-borne data across diverse environments, such as lakes, rivers, seas, or lagoons. Despite being a common option for bathymetry mapping, the use of satellite imagery faces challenges due to the complex inherent optical properties of water bodies (e.g., turbid water), satellite spatial resolution limitations, and constraints in the performance of retrieval models. This study focuses on advancing the remote sensing based method by harnessing the non-linear learning capabilities of the machine learning (ML) model, employing advanced feature selection through a meta-heuristic algorithm, and using image extraction techniques (i.e., band ratio, gray scale morphological operation, and morphological multi-scale decomposition). Herein, we validate the predictive capabilities of six ML models: Random Forest (RF), Support Vector Machine (SVM), CatBoost (CB), Extreme Gradient Boost (XGB), Light Gradient Boosting Machine (LGBM), and KTBoost (KTB) models, both with and without the application of meta-heuristic optimization (i.e., Dragon Fly, Particle Swarm Optimization, and Grey Wolf Optimization), to accurately ascertain water depth. This is achieved using a diverse input dataset derived from multi-spectral Landsat 9 imagery captured on a cloud-free day (19 September 2023) in a shallow, turbid lagoon. Our findings indicate the superior performance of LGBM coupled with Particle Swamp Optimization (R2 = 0.908, RMSE = 0.31 m), affirming the consistency and reliability of the feature extraction and selection-based framework, while offering novel insights into the expansion of bathymetric mapping in complex aquatic environments. [ABSTRACT FROM AUTHOR]

Document classification is becoming increasingly essential for the vast number of documents available in digital libraries, emails, the Internet, etc. Textual records frequently contain non-discriminative (noisy and irrelevant) terms that are also high-dimensional, resulting in higher computing costs and poorer learning performance in Text Classification (TC). Feature selection (FS), which tries to discover discriminate terms or features from the textual data, is one of the most effective tasks for this issue. This paper introduces a novel multi-stage term-weighting scheme-based FS model designed for the single-label TC system to obtain the optimal set of features. We have also developed a hybrid deep learning fine-tuning network based on Bidirectional Long Short-Term Memory (BiLSTM) and Convolutional Neural Network (CNN) for the classification stage. The FS approach is worked on two-stage criteria. The filter model is used in the first stage, and the multi-objective wrapper model, an upgraded version of the Whale Optimization Algorithm (WOA) with Particle Swarm Optimization (PSO), is used in the second stage. The objective function in the above wrapper model is based on a tri-objective principle. It uses the Pareto front technique to discover the optimal set of features. Here in the wrapper model, a novel selection strategy has been introduced to select the whale instead of the random whale. The proposed work is evaluated on four popular benchmark text corpora, of which two are binary class, and two are multi-class. The suggested FS technique is compared against classic Machine Learning (ML) and deep learning classifiers. The results of the experiments reveal that the recommended FS technique is more effective in obtaining better results than the other results. [ABSTRACT FROM AUTHOR]

Securing Unmanned Aerial Vehicle (UAV) systems is vital to safeguard the processes involved in operating the drones. This involves the execution of robust communication encryption processes to defend the data exchanged between the UAVs and ground control stations. Intrusion detection, powered by Deep Learning (DL) techniques such as Convolutional Neural Networks (CNN), allows the classification and identification of potential attacks or illegal objects in the operational region of the drone, thus distinguishing them from the routine basics. The current research work offers a new Hybrid Arithmetic Optimizer Algorithm with DL method for Secure Unmanned Aerial Vehicle Network (HAOADL-UAVN) model. The purpose of the proposed HAOADL-UAVN technique is to secure the communication that occurs in UAV networks via threat detection. At the primary level, the network data is normalized through min-max normalization approach in order to scale the input dataset into a useful format. The HAOA is used to select a set of optimal features. Next, the security is attained via Deep Belief Network Autoencoder (DBN-AE)-based threat detection. At last, the hyperparameter choice of the DBN-AE method is implemented using the Seagull Optimization Algorithm (SOA). A huge array of simulations was conducted using the benchmark datasets to demonstrate the improved performance of the proposed HAOADL-UAVN algorithm. The comprehensive results underline the supremacy of the HAOADL-UAVN methodology under distinct evaluation metrics. [ABSTRACT FROM AUTHOR]

An IDS (Intrusion Detection System) is essential for network security experts, as it allows one to identify and respond to abnormal traffic present in a network. An IDS can be utilized for evaluating the various types of malicious attacks. Hence, detecting intrusions has become a significant research area in the contemporary era, especially with the evolution of technologies. With the progress of ML (Machine Learning)-based algorithms, researchers have striven to perform optimal ID. However, most of these studies lag in accordance with their accuracy rate. Thus, to attain a high accuracy rate in ID, the present study proposes ML-based meta-heuristic algorithms, as these approaches possess innate merits of determining near-optimal solutions in limited time and are capable of dealing with multi-dimensional data. The study proposes OWSA (Optimal Whale Sine Algorithm) for selecting suitable and relevant features. With an exclusive optimization process using the SCA (Sine Cosine Algorithm), this study proposes to combine SCA with WOA (Whale Optimization Algorithm) for mitigating the demerits of both, with its hybridization thereby achieving OWSA. Following this, AWRF (Artificial Neural Network Weighted Random Forest) is proposed for classification. The main intention of this process is to propose a weight-updating process for discrete trees in the RF model. The proposed approach is motivated by avoiding overfitting and attaining stability and flexibility. This approach is assessed with regard to performance via a comparative analysis, so as to uncover the best performance of this proposed technique in ID. [ABSTRACT FROM AUTHOR]

Feature selection is an important component of the machine learning domain, which selects the ideal subset of characteristics relative to the target data by omitting irrelevant data. For a given number of features, there are 2n possible feature subsets, making it challenging to select the optimal set of features from a dataset via conventional feature selection approaches. We opted to investigate glaucoma infection since the number of individuals with this disease is rising quickly around the world. The goal of this study is to use the feature set (features derived from fundus images of benchmark datasets) to classify images into two classes (infected and normal) and to select the fewest features (feature selection) to achieve the best performance on various efficiency measuring metrics. In light of this, the paper implements and recommends a metaheuristics-based technique for feature selection based on emperor penguin optimization, bacterial foraging optimization, and proposes their hybrid algorithm. From the retinal fundus benchmark images, a total of 36 features were extracted. The proposed technique for selecting features minimizes the number of features while improving classification accuracy. Six machine learning classifiers classify on the basis of a smaller subset of features provided by these three optimization techniques. In addition to the execution time, eight statistically based performance metrics are calculated. The hybrid optimization technique combined with random forest achieves the highest accuracy, up to 0.95410. Because the proposed medical decision support system is effective and ensures trustworthy decision-making for glaucoma screening, it might be utilized by medical practitioners as a second opinion tool, as well as assist overworked expert ophthalmologists and prevent individuals from losing their eyesight. [ABSTRACT FROM AUTHOR]

The selection of important factors in machine learning-based susceptibility assessments is crucial to obtain reliable susceptibility results. In this study, metaheuristic optimization and feature selection techniques were applied to identify the most important input parameters for mapping debris flow susceptibility in the southern mountain area of Chengde City in Hebei Province, China, by using machine learning algorithms. In total, 133 historical debris flow records and 16 related factors were selected. The support vector machine (SVM) was first used as the base classifier, and then a hybrid model was introduced by a two-step process. First, the particle swarm optimization (PSO) algorithm was employed to select the SVM model hyperparameters. Second, two feature selection algorithms, namely principal component analysis (PCA) and PSO, were integrated into the PSO-based SVM model, which generated the PCA-PSO-SVM and FS-PSO-SVM models, respectively. Three statistical metrics (accuracy, recall, and specificity) and the area under the receiver operating characteristic curve (AUC) were employed to evaluate and validate the performance of the models. The results indicated that the feature selection-based models exhibited the best performance, followed by the PSO-based SVM and SVM models. Moreover, the performance of the FS-PSO-SVM model was better than that of the PCA-PSO-SVM model, showing the highest AUC, accuracy, recall, and specificity values in both the training and testing processes. It was found that the selection of optimal features is crucial to improving the reliability of debris flow susceptibility assessment results. Moreover, the PSO algorithm was found to be not only an effective tool for hyperparameter optimization, but also a useful feature selection algorithm to improve prediction accuracies of debris flow susceptibility by using machine learning algorithms. The high and very high debris flow susceptibility zone appropriately covers 38.01% of the study area, where debris flow may occur under intensive human activities and heavy rainfall events. [ABSTRACT FROM AUTHOR]

In today's competitive market, predicting clients' behavior is crucial for businesses to meet their needs and prevent them from being attracted by competitors. This is especially important in industries like telecommunications, where the cost of acquiring new customers exceeds retaining existing ones. To achieve this, companies employ Customer Churn Prediction approaches to identify potential customer attrition and develop retention plans. Machine learning models are highly effective in identifying such customers; however, there is a need for more effective techniques to handle class imbalance in churn datasets and enhance prediction accuracy in complex churn prediction datasets. To address these challenges, we propose a novel two-level stacking-mode ensemble learning model that utilizes the Whale Optimization Algorithm for feature selection and hyper-parameter optimization. We also introduce a method combining K-member clustering and Whale Optimization to effectively handle class imbalance in churn datasets. Extensive experiments conducted on well-known datasets, along with comparisons to other machine learning models and existing churn prediction methods, demonstrate the superiority of the proposed approach. [ABSTRACT FROM AUTHOR]

The rapid development of data science has led to the emergence of high-dimensional datasets in machine learning. The curse of dimensionality is a significant problem caused by high-dimensional data with a small sample size. This paper proposes a novel hybrid binary dragonfly algorithm (HBDFA) in which a distance-based similarity evaluation algorithm is embedded before the dragonfly algorithm (DA) searching behavior to select the most discriminating features. The two-step feature selection mechanism of HBDFA enables the method to explore the feature space reduced by the distance-based similarity evaluation algorithm. The model was evaluated on two datasets. The first dataset contained 200 reports from 4 evenly distributed categories of Daily Mail Online: COVID-19, economy, science, and sports. The second dataset was the publicly available Spam dataset. The proposed model is compared with binary versions of four popular metaheuristic algorithms. The model achieved an accuracy rate of 96.75% by reducing 66.5% of the top 100 features determined on the first dataset. Results on the Spam dataset reveal that HBDFA gives the best classification results with over 95% accuracy. The experimental results show the superiority of HBDFA in searching high-dimensional data, improving classification results, and reducing the number of selected features. [ABSTRACT FROM AUTHOR]

Feature selection is becoming a relevant problem within the field of machine learning. The feature selection problem focuses on the selection of the small, necessary, and sufficient subset of features that represent the general set of features, eliminating redundant and irrelevant information. Given the importance of the topic, in recent years there has been a boom in the study of the problem, generating a large number of related investigations. Given this, this work analyzes 161 articles published between 2019 and 2023 (20 April 2023), emphasizing the formulation of the problem and performance measures, and proposing classifications for the objective functions and evaluation metrics. Furthermore, an in-depth description and analysis of metaheuristics, benchmark datasets, and practical real-world applications are presented. Finally, in light of recent advances, this review paper provides future research opportunities. [ABSTRACT FROM AUTHOR]

The dropout rate in underdeveloped and emerging countries is a pressing social issue, as highlighted by studies conducted by The Organization for Economic Co-operation and Development. This study compares five feature selection techniques to address this challenge and improve the automatic detection of dropout risk. The methodological design involves three distinct phases: data preparation, feature selection, and model evaluation utilizing machine learning algorithms. The results demonstrate that (1) the top features identified by feature selection techniques, i.e., those constructed through feature engineering, proved to be among the most effective in classifying student dropout; (2) the F-score of the best model increased by 5% with feature selection techniques; and (3) depending on the type of feature selection, the performance of the machine learning algorithm can vary, potentially increasing or decreasing based on the sensitivity of features with higher noise. At the same time, metaheuristic algorithms demonstrated significant precision improvements, but there was a risk of increasing errors and reducing recall.

• A novel integrated framework for the diagnosis of breast cancer patients is proposed. • A scheme combining feature selection and hyperparameter optimization is proposed to improve the prediction accuracy of the model. • Compared with the existing models, the proposed model has higher performance and generalization ability. In this paper, a new integrated scheme is proposed to accurately predict breast cancer, help doctors make early diagnosis and treatment plans, and improve the prognosis of patients. We selects five mainstream machine learning models: support vector machine (SVM), artificial neural network (ANN), random forest (RF), extreme gradient boosting (XGBoost) and adaptive boosting (AdaBoost). The Wisconsin Breast Cancer Database (WBCD) and Wisconsin Diagnostic Breast Cancer (WDBC) are used as datasets to investigate the predictive performance of these single and ensemble models. Then, we use multiple linear regression method for feature selection (FS), the experimental results show that the change of feature subset will significantly affect the performance of the model. The recall and f1-score of the five models are improved by 1.19% and 0.84% on average. After that, we apply whale optimization algorithm (WOA) to optimize the hyperparameters of the model to improve their prediction performance. In the best-case scenario, the model demonstrated improvements of 1.02% in accuracy and 1.82% in precision. In addition, we ensemble these models by stacking, investigate the performance changes of the ensemble model when different models are used as meta learners. Finally, the FS-WOA-Stacking model achieves 99.56% accuracy on WBCD and 99.65% accuracy on WDBC. Compared with the existing breast cancer prediction models, the performance of the proposed model is at an excellent level. [ABSTRACT FROM AUTHOR]

Optimization problems can be seen in numerous fields of practical studies. One area making waves in the application of optimization methods is data mining in machine learning. An important preprocessing technique of data mining where irrelevant variables are discarded from the datasets and holding onto variables with important information is referred to as feature selection (FS). FS is critical to tackling the 'curse of dimensionality' by reducing the number of features, minimizing computational expensiveness and maximizing the accuracy of the machine learning models. Swarm Intelligence (SI)-based meta-heuristic algorithms (MAs) have been widely employed to solve several optimization problems like FS. However, common drawbacks identified with these algorithms include getting trapped in local optima, especially in situations where the search space is large (high dimensional space). This study proposes a new hybrid SI-based MA called Particle Swarm-guided Bald Eagle Search (PS-BES). The algorithm utilizes the speed of Particle Swarm to guide Bald Eagles in their search to ensure a smooth transition of the algorithm from exploration to exploitation. Additionally, we introduce the Attack-Retreat-Surrender technique, a new local-optima escape technique to enhance the balance between diversification and intensification of PS-BES. To establish the outstanding performance of the proposed algorithm, PS-BES is comprehensively analyzed utilizing 26 Benchmark functions. Further, the practicality of PS-BES is highlighted by its binary version for feature selection and evaluated using 27 classification datasets from the UCI repository. The results prove the overall superiority of PS-BES and bPS-BES as opposed the 10 state-of-the-art algorithms employed in the study. • Metaheuristic methods have proven to effectively tackle optimization problems. • Wrapper-based feature selection technique efficiently evaluate important subspaces. • The Attack-Retreat-Surrender technique fosters the robust search strategy in PS-BES. • PS-BES and bPS-BES show outstanding performance over the competing methods. [ABSTRACT FROM AUTHOR]

Computational and technological advancements have led to an increase in data generation and storage capacity. Many annotated datasets have been used to train machine learning models for predictive tasks. Feature selection (FS) is a combinatorial binary optimization problem that arises from a need to reduce dataset dimensionality by finding the subset of features with maximum predictive accuracy. While different methodologies have been proposed, metaheuristics adapted to binary optimization have proven to be reliable and efficient techniques for FS. This paper applies the first and unique population-trajectory metaheuristic, the Lichtenberg algorithm (LA), and enhances it with a Fibonacci sequence to improve its exploration capabilities in FS. Substituting the random scales that controls the Lichtenberg figures' size and the population distribution in the original version by a sequence based on the golden ratio, a new optimal exploration–exploitation LF's size decay is presented. The new few hyperparameters golden Lichtenberg algorithm (GLA), LA, and eight other popular metaheuristics are then equipped with the v-shaped transfer function and associated with the K-nearest neighbor classifier in the search of the optimized feature subsets through a double cross-validation experiment method on 15 UCI machine learning repository datasets. The binary GLA selected reduced subsets of features, leading to the best predictive accuracy and fitness values at the lowest computational cost. [ABSTRACT FROM AUTHOR]

Alzheimer's disease (AD) is a chronic disorder characterized by progressive cognitive dysfunctions and memory loss. Electroencephalography (EEG) is a non-invasive tool to detect AD using machine learning models and signal-derived features. Feature selection enhances AD detection performance, reduces training and testing time, and builds simpler models. Metaheuristic algorithms (MHAs) have successfully selected optimum features in several classification tasks and can be valuable in enhancing AD detection. In this work, an AD detection model is proposed based on a systematic investigation of adaptive signal decomposition techniques and MHAs. The four techniques comprising empirical mode decomposition, variational mode decomposition, discrete wavelet transform, and Low-Complexity Orthogonal Wavelet Filter Banks (LCOWFBs) decompose EEG signals into subbands. Thirty-four features based on signal, temporal, spectral, and entropy analysis are extracted for each subband. The salient features are selected using seven different MHAs: particle swarm algorithm, salp swarm algorithm, reptile search (RSA), Grey Wolf, dwarf mongoose optimization, snake optimizer, and Fick's law algorithm (FLA). The two publicly available AD EEG datasets validated the performance of the proposed framework. LCOWFBs achieved the highest binary classification accuracy of 99.72% using 15 salient features selected by RSA and multi-class accuracy of 88.92% with seven salient features selected by FLA. The present method investigates brain regions that are affected by AD. The results indicate that the combination of LCOWFBs and RSA achieved maximum AD detection performance for both datasets with shorter computational time than earlier reported methods in the literature. The present model can be extended to identify chronic disorders, including hypertension, sleep disorders, and Parkinson's disease. • Developing an Metaheuristic Algorithms-based optimal feature selection model for the detection of AD and MCI from normal. • Presenting the signal processing decomposition techniques EMD, VMD, DWT, and LCOWFBs for feature extraction in bandwidth-duration domain. • Computing temporal, spectral, and entropy-based thirty-five features to explore the significant brain regions. • Evaluating classification performance on two publically available datasets with tuned hyper-parameters of MHAs methods for AD and MCI detection. • Enhancing AD detection performance by feature selection of present model and comparing research findings with previous techniques. [ABSTRACT FROM AUTHOR]

• An intelligent framework is developed to frequently update set points in the grinding circuit, aiming to maximise SAG mill throughput. • The prediction model can be frequently retrained using data from process sensors and updated to adapt to changing conditions of the grinding process. This dynamic approach distinguishes it from static modelling methods, which may experience reduced accuracy over time. • CatBoost is the most accurate predictor of SAG mill throughput among 17 machine learning models evaluated. Differential Evolution outperforms other optimisation algorithms. In mineral processing plants, grinding is a crucial step, accounting for approximately 50% of the total mineral processing costs. Semi-autogenous grinding (SAG) mills are extensively employed in the grinding circuit of mineral processing plants. Maximising SAG mill throughput is of significant importance considering its profound financial outcomes. However, the optimum process parameter setting aimed at achieving maximum mill throughput remains an uninvestigated domain in prior research. This study introduces an intelligent framework leveraging expert knowledge, machine learning techniques and evolutionary algorithms to address this research need. In this study, an extensive industrial dataset comprising 36,743 records is utilised and relevant features are selected based on the insights of industry experts. Following the removal of erroneous data, an evaluation of 17 machine learning models is undertaken to identify the most accurate predictive model. To improve the performance of the model, feature selection and outlier detection are executed. The resultant optimal model, trained with refined features, serves as the objective function within three distinct evolutionary algorithms. These algorithms are employed to identify parameter configurations that maximise SAG mill throughput while adhering to the working limits of input parameters as constraints. Notably, analysis revealed that CatBoost, as an ensemble model, stands out as the most accurate predictor. Furthermore, differential evolution emerges as the preferred optimisation algorithm, exhibiting superior performance in both achieving the highest mill throughput predictions and ensuring robustness in predictions, surpassing alternative methods. [ABSTRACT FROM AUTHOR]

• A feature selection method based on binary dandelion algorithm is proposed. • The algorithm applies a sine-cosine operator and a restart strategy. • Mutual information and quick bit mutation improve the performance of the algorithm. • The algorithm performs well on datasets with larger dimensions. Feature selection (FS) is an important data preprocessing technology for machine learning and data mining. Metaheuristic algorithm (MH) has been widely used in feature selection because of its powerful search function. This paper presents an improved Binary Dandelion Algorithm using Sine Cosine operator and Restart strategy (SCRBDA) for feature selection. First, the sine cosine operator is used in the radius formula of the core dandelions (CD), which significantly enhances the ability of algorithm development and exploration. Secondly, the algorithm uses a restart strategy to increase its ability to get rid of local optimum. Thirdly, mutual information is used to guide the generation of some dandelions, which pays more attention to the correlation between the selected features and categories. Finally, quick bit mutation is used as the mutation strategy to improve the diversity of the population. The SCRBDA proposed in this paper was tested on 18 datasets of different sizes from UCI machine learning database. The SCRBDA was compared with 8 other classical feature selection algorithms, and the performance of the proposed algorithm was evaluated through feature subset size, classification accuracy, fitness value, and F1-score. The experimental results show that SCRBDA achieves the best performance, which has stronger feature reduction ability and achieves better overall performance on most datasets. Especially on large-scale datasets, SCRBDA can obtain extremely smaller feature subsets while maintaining much higher classification accuracy, and satisfactory F1-score. [ABSTRACT FROM AUTHOR]

[Display omitted] Feature selection (FS) is a crucial process that aims to remove unnecessary features from datasets. It plays a role in data mining and machine learning (ML) by reducing the risk associated with high-dimensional datasets. FS is considered a challenging problem that is difficult to solve efficiently due to its combinatorial nature. As the size of the problem increases, the computation time also grows. Recently, researchers have focused on metaheuristic FS algorithms specifically designed for high-dimensional datasets. Therefore, this article proposes a powerful metaheuristic algorithm called Binary Enhanced Golden Jackal Optimization (BEGJO), which is an improved version of the recently published Golden Jackal Optimization (GJO) algorithm. The original GJO algorithm faces challenges when dealing with high-dimensional FS problems, as it tends to get trapped in local optima. To address this issue, various enhancement strategies are employed to improve the efficiency of GJO. The proposed BEGJO algorithm utilizes Copula Entropy (CE) to reduce the dimensionality of high-dimensional FS problems while maintaining high classification accuracy using the K-Nearest Neighbour (K-NN) classifier. Additionally, four enhancement strategies are incorporated to enhance the exploration and exploitation capabilities of the fundamental GJO algorithm. The BEGJO algorithm is transformed into its binary form using the sigmoid transfer function, aligning it with the nature of the FS problem. It is then tested on various high-dimensional benchmark datasets. The effectiveness of BEGJO is evaluated by comparing it with well-known algorithms in terms of classification accuracy, feature dimension, and processing time. BEGJO outperforms other algorithms in terms of classification accuracy and feature dimension and ranks up to fourth in terms of processing time. Furthermore, the advantageous use of CE is demonstrated by comparing the performance of the proposed algorithm with traditional FS algorithms. Statistical evaluations are conducted to further validate the effectiveness and superiority of the proposed algorithm. The results confirm that BEGJO is an effective solution for high-dimensional FS problems. [ABSTRACT FROM AUTHOR]

Diabetes is a dreaded disease that can be identified by elevated blood glucose levels in the blood, and undiagnosed diabetes can cause a host of related complications, such as retinopathy and nephropathy. In terms of type, the main categories are type 1 diabetes (T1DM), type 2 diabetes (T2DM) and gestational diabetes mellitus (GDM). Machine learning models and metaheuristic optimization algorithms can play an important role in the early detection, diagnosis and treatment of this disease. To this end, we propose AHDHS-Stacking, an ensemble learning framework for diabetes mellitus classification and diagnosis that is based on the harmony search (HS) algorithm and stacking and includes two stages of feature selection and optimization of base-learner combinations. To improve the model's overall performance, the average performance of all base learners is used as the feature selection target, and an adaptive hyperparameter strategy is used to accelerate the iterative process. HS is then used to optimize to find the best combination of base learners, which improves model performance while reducing complexity. Following that, we conducted experiments on the Pima Indians Diabetes (PID) dataset and the Chinese and Western Medicine Diabetes (CWMD) dataset, achieving accuracy of 93.09%, precision of 93.22%, recall of 91.60% , F-measure of 92.25%, and MCC of 84.79% on PID dataset, which is better than all benchmark models and validated the model's validity. CWMD dataset experimental results showed that AHDHS-Stacking screened for key features such as age, gender, urinary glucose, fasting glucose, BMI and cholesterol, and can be used as a practical and accurate method for early diabetes prediction. [ABSTRACT FROM AUTHOR]