Feature Weighting Research Articles

Hybrid model of feature-driven modular neural network-based grasshopper optimization algorithm for diabetic retinopathy classification using fundus images.

Diabetic retinopathy (DR) is a progressive condition that can lead to blindness if undiagnosed or untreated. Automatic systems for DR prediction using fundus images have been developed, but challenges like variable illumination, overfitting, small datasets, poor feature learning, high computational complexity, and suboptimal feature weighting persist. To address these, a hybrid model called the modular neural network with grasshopper optimization algorithm (MNN-GOA) is proposed. This model integrates neural network capabilities with the grasshopper optimization algorithm (GOA) to enhance feature selection and classification accuracy. It begins with preprocessing to improve image quality, followed by data augmentation and histogram-based segmentation to focus on critical regions. Features are extracted using techniques like histogram of oriented gradients (HOG), scale-invariant feature transform (SIFT), color features, and mutual information (MI). GOA optimizes feature weights, balancing exploration and exploitation, while reducing computational complexity. The model integrates features from ground truth and original images to predict DR stages accurately. Achieving performance metrics of accuracy (98.8%), specificity (97.6%), sensitivity (96.8%), precision (96.4%), and F1 score (96.2%), the MNN-GOA model was validated on four datasets like DIARETDB1, DDR, APTOS 2019, and EyePACS and outperformed existing methods, proving to be a robust and efficient solution for DR classification and severity prediction.

Multifaceted approach for anticipating learner performance using parameter weightage and ensemble algorithm fusion

Anticipating student performance has garnered significant attention in education research for offering early insights that enable timely interventions and personalized support, ultimately improving student success and retention rates. This research focuses on enhancing the accuracy and efficiency of student performance prediction models by employing a hybrid ensemble framework that integrates weighted feature selection with meta-learner-based approaches. A weighted feature selection method was employed to prioritize the most influential of the 23 parameters in the dataset, enhancing prediction accuracy while reducing the computational burden. These parameters were then used to build a hybrid ensemble model by combining base learners with meta-learners, systematically tuned using hyperparameter optimization. This approach aimed to further improve prediction accuracy by fusing multiple base learners, leveraging the strengths of different algorithms for more accurate predictions. The proposed hybrid model was validated across different features selected based on feature importance using random forest (RF). An accuracy of 98.38% was achieved when all 23 features were considered and an accuracy of 97.13 % was achieved when the top 10 features were used. The research highlights the significance of early prediction for prompt intervention and demonstrates how feature weighting can boost model efficacy.

AL-FEW: An enhanced approach for optimized query examples through feature weighting in active learning

AL-FEW: An enhanced approach for optimized query examples through feature weighting in active learning

SPDC-YOLO: An Efficient Small Target Detection Network Based on Improved YOLOv8 for Drone Aerial Image

Target detection in UAV images is of great significance in fields such as traffic safety, emergency rescue, and environmental monitoring. However, images captured by UAVs usually have multi-scale features, complex backgrounds, uneven illumination, and low target resolution, which makes target detection in UAV images very challenging. To tackle these challenges, this paper introduces SPDC-YOLO, a novel model built upon YOLOv8. In the backbone, the model eliminates the last C2f module and the final downsampling module, thus avoiding the loss of small target features. In the neck, this paper proposes a novel feature pyramid, SPC-FPN, which employs the SBA (Selective Boundary Aggregation) module to fuse features from two distinct scales. In the head, the P5 detection head is eliminated, and a new detection head, Dyhead-DCNv4, is proposed, replacing DCNv2 in the original Dyhead with DCNv4 and utilizing three attention mechanisms for dynamic feature weighting. In addition, the model uses the CGB (Context Guided Block) module for downsampling, which can learn and fuse local features with surrounding contextual information, and the PPA (Parallelized Patch-Aware Attention) module replacing the original C2f module to further improve feature expression capability. Finally, SPDC-YOLO adopts EIoU as the loss function to optimize target localization accuracy. On the public dataset VisDrone2019, the experimental results show that SPDC-YOLO improves mAP50 by 3.4% compared to YOLOv8n while reducing the parameters count by 1.03 M. Compared with other related methods, SPDC-YOLO demonstrates better performance.

CMADRL: cross-modal attention based deep reinforcement learning for mobile robot’s obstacle avoidance

Abstract The multimodal perception is very crucial for mobile robots to achieve safe autonomous
steering in complex and changing environments without colliding with obstacles. In spite of the fact that a rich environmental information can be provided by the multimodal data, how to integrate the complementary features from different sources effectively remains an open issue. In this paper, a simple yet effective multimodal deep reinforcement learning (DRL) method, CMADRL, by fusing two input modalities such as depth images and pseudo-LiDAR data generated by RGB-D cameras, is proposed to improve the performance of single sensor in environmental perception tasks. And the Convolutional LSTM Network (ConvLSTM) layers are also adopted to extract spatiotemporal features and capture temporal relationships in consecutive time steps, which allow the robot to predict dynamic changes of the environment. In addition, A crossmodal fusion module is designed to optimize the multimodal data fusion scheme, with dynamic weighting and merging of features from different modalities, allowing the agent to focus more reasonably on the current state and improving the accuracy and efficiency of the obstacle avoidance strategy. The experimental results show that the proposed approach achieved significant performance in cumulative rewards, convergence speed, and success rate. Additionally, tests in both simulated and real-world environments further verify that collisions are effectively avoided using only a low-cost RGB-D camera, demonstrating the method’s strong generalization capabilities.

Classification of Royal Delicious Apples using Hybrid Feature Selection and Feature Weighting Method Based on SVM Classifier

Fruit safety is a critical component of the global economy, particularly within the agricultural sector. There has been a recent surge in the incidence of diseases affecting fruits, leading to economic setbacks in agriculture worldwide. Conventional manual assessment methods are laborious, prompting the exploration of automated computerized techniques for evaluating fruit quality. This research presents a novel method for assessing the quality of golden delicious apples. A dataset comprising 1256 apple images was gathered under controlled conditions. Afterward Feature extraction focuses on texture features like LBP, GLCM, GLDM, DTF, and Gabor features, color features, and shape and size features. A total of 18654 features are extracted and normalized using z-score. A hybrid method for feature selection and weighting involves the mRMR algorithm to eliminate redundant features and the Sine Cosine Optimization Algorithm for feature weighting, enhancing classification performance. The SVM machine learning technique, augmented with optimized features, yielded a 10.53% improvement in accuracy compared to SVM alone. Validation against state-of-the-art methods using Friedman's mean rank test underscored the statistical significance of this approach across various metrics.

Image recognition technology for bituminous concrete reservoir panel cracks based on deep learning.

Detecting cracks in asphalt concrete slabs is challenging due to environmental factors like lighting changes, surface reflections, and weather conditions, which affect image quality and crack detection accuracy. This study introduces a novel deep learning-based anomaly model for effective crack detection. A large dataset of panel images was collected and processed using denoising, standardization, and data augmentation techniques, with crack areas labeled via LabelImg software. The core model is an improved Xception network, enhanced with an adaptive activation function, dynamic attention mechanism, and multi-level residual connections. These innovations optimize feature extraction, enhance feature weighting, and improve information transmission, significantly boosting accuracy and robustness. The improved model achieves a 97.6% accuracy and a Matthews correlation coefficient of 0.98, remaining stable under varying lighting conditions. This method not only provides a fresh approach to crack detection but also greatly enhances detection efficiency.

A novel case-based reasoning system for explainable lung cancer diagnosis.

A novel case-based reasoning system for explainable lung cancer diagnosis.

An optimized ensemble grey wolf-based pipeline for monkeypox diagnosis

As the world recovered from the coronavirus, the emergence of the monkeypox virus signaled a potential new pandemic, highlighting the need for faster and more efficient diagnostic methods. This study introduces a hybrid architecture for automatic monkeypox diagnosis by leveraging a modified grey wolf optimization model for effective feature selection and weighting. Additionally, the system uses an ensemble of classifiers, incorporating confusion based voting scheme to combine salient data features. Evaluation on public data sets, at various of training samples percentages, showed that the proposed strategy achieves promising performance. Namely, the system yielded an overall accuracy of 98.91% with testing run time of 5.5 seconds, while using machine classifiers with small number of hyper-parameters. Additional experimental comparison reveals superior performance of the proposed system over literature approaches using various metrics. Statistical analysis also confirmed that the proposed AMDS outperformed other models after running 50 times. Finally, the generalizability of the proposed model is evaluated by testing its performance on external data sets for monkeypox and COVID-19. Our model achieved an overall diagnostic accuracy of 98.00% and 99.00% on external COVID and monkeypox data sets, respectively.

Multi-view human point cloud registration method with overlapping regions semantic constraints and feature weighting

Multi-view human point cloud registration method with overlapping regions semantic constraints and feature weighting

Ultrasound-Guided Vacuum-Assisted Excision (VAE) in Breast Lesion Management: An Experimental Comparative Study of Two Different VAE Devices Across Various Aspiration Levels and Window Sizes.

Background/Objectives: Vacuum-assisted excision (VAE) is a minimally invasive technique for breast tumor treatment, offering precision, comfort, and quick recovery. It is widely used for benign breast lesions and is playing an increasingly important role in the therapeutic management of non-surgical patients or patients who refuse surgery. Optimal outcomes require an understanding of device features to tailor treatment to each lesion. The Mammotome® Elite 10G operates in a fixed mode, while the Mammotome® Revolve EX 8G offers multiple aspiration levels and aperture windows for greater versatility. This study analyzed the specimen features (weight and length), comparing the weight obtained from two different VAE systems to aid the appropriate selection of a device based on the clinical setting. It also determined the number of specimens needed to achieve the 4 g diagnostic threshold. Methods: The Mammotome® Elite 10G and the Mammotome® Revolve EX were evaluated under controlled conditions. For Mammotome® Revolve EX, combinations of five aspiration levels and three aperture lengths (12 mm, 18 mm, and 25 mm) were tested. Twelve samples were collected from a chicken breast phantom for each setting. Specimen weights and the minimum excisions required to reach the 4 g threshold were analyzed. Results: The mean weight per sample for the Mammotome® Elite 10G was 0.16 ± 0.04 g. For the Mammotome® Revolve EX, the weights increased with aperture size and aspiration level, ranging from a minimum of 0.132 ± 0.028 g (a window length of 12 mm and aspiration level 1) to a maximum of 0.407 ± 0.055 g (a window length of 25 mm and aspiration level 5). The 25 mm window at aspiration level 5 achieved the 4 g threshold in as few as 10 samples. By comparison, the Mammotome® Elite required up to 26 samples. Conclusions: Compared to the Mammotome Elite, Mammotome® Revolve EX offers superior versatility and efficiency, reducing patient discomfort by minimizing the required samples. Its technical advantages make it a valuable tool for both diagnostic and therapeutic applications.

Graph Neural Network and LSTM Integration for Enhanced Multi-Label Style Classification of Piano Sonatas.

In the field of musicology, the automatic style classification of compositions such as piano sonatas presents significant challenges because of their intricate structural and temporal characteristics. Traditional approaches often fail to capture the nuanced relationships inherent in musical works. This paper addresses the limitations of traditional neural networks in piano sonata style classification and feature extraction by proposing a novel integration of graph convolutional neural networks (GCNs), graph attention networks (GATs), and Long Short-Term Memory (LSTM) networks to conduct the automatic multi-label classification of piano sonatas. Specifically, the method combines the graph convolution operations of GCNs, the attention mechanism of GATs, and the gating mechanism of LSTMs to perform the graph structure representation, feature extraction, allocation weighting, and coding of time-dependent features of music data layer by layer. The aim is to optimize the representation of the structural and temporal features of musical elements, as well as the dependence between discovery features, so as to improve classification performance. In addition, we utilize MIDI files of several piano sonatas to construct a dataset, spanning the 17th to the 19th centuries (i.e., the late Baroque, Classical, and Romantic periods). The experimental results demonstrate that the proposed method effectively improves the accuracy of style classification by 15% over baseline schemes.

Klasifikasi Sentimen Masyarakat di Twitter terhadap Puan Maharani dengan Metode Modified K-Nearest Neighbor

This study aims to address the challenges in classifying sentiment on Twitter regarding Puan Maharani by implementing the Modified K-Nearest Neighbor (MK-NN) method, supplemented with feature weighting and feature selection techniques. This method is designed to improve accuracy by assigning higher weights to important features and reducing data dimensions to avoid overfitting. Data is collected using a crawling technique on Indonesian-language tweets, which are then manually labeled and processed through a preprocessing stage. The testing results using the modified K-Nearest Neighbor (MK-NN) method with confusion matrices show the model's performance at three different values of K (3, 5, and 7) and data ratios of 90:10, 80:20, and 70:30. With a 90:10 data ratio and K=3, the method achieved the highest accuracy of 89.0%. These results indicate that the combination of MK-NN and related techniques is highly effective in sentiment classification, offering an innovative solution to the limitations of conventional methods. These findings have potential applications in public opinion analysis, particularly for supporting data-driven strategic decision-making.

Speech Emotion Recognition Model Based on Joint Modeling of Discrete and Dimensional Emotion Representation

This paper introduces a novel joint model architecture for Speech Emotion Recognition (SER) that integrates both discrete and dimensional emotional representations, allowing for the simultaneous training of classification and regression tasks to improve the comprehensiveness and interpretability of emotion recognition. By employing a joint loss function that combines categorical and regression losses, the model ensures balanced optimization across tasks, with experiments exploring various weighting schemes using a tunable parameter to adjust task importance. Two adaptive weight balancing schemes, Dynamic Weighting and Joint Weighting, further enhance performance by dynamically adjusting task weights based on optimization progress and ensuring balanced emotion representation during backpropagation. The architecture employs parallel feature extraction through independent encoders, designed to capture unique features from multiple modalities, including Mel-frequency Cepstral Coefficients (MFCC), Short-term Features (STF), Mel-spectrograms, and raw audio signals. Additionally, pre-trained models such as Wav2Vec 2.0 and HuBERT are integrated to leverage their robust latent features. The inclusion of self-attention and co-attention mechanisms allows the model to capture relationships between input modalities and interdependencies among features, further improving its interpretability and integration capabilities. Experiments conducted on the IEMOCAP dataset using a leave-one-subject-out approach demonstrate the model’s effectiveness, with results showing a 1–2% accuracy improvement over classification-only models. The optimal configuration, incorporating the joint architecture, dynamic weighting, and parallel processing of multimodal features, achieves a weighted accuracy of 72.66%, an unweighted accuracy of 73.22%, and a mean Concordance Correlation Coefficient (CCC) of 0.3717. These results validate the effectiveness of the proposed joint model architecture and adaptive balancing weight schemes in improving SER performance.

Lightweight YOLOv8 for tongue teeth marks and fissures detection based on C2f_DCNv3

This paper propose a significantly enhanced YOLOv8 model specifically designed for detecting tongue fissures and teeth marks in Traditional Chinese Medicine (TCM) diagnostic images. By integrating the C2f_DCNv3 module, which incorporates Deformable Convolutions (DCN), replace the original C2f module, enabling the model to exhibit exceptional adaptability to intricate and irregular features, such as fine fissures and teeth marks. Furthermore, the introduction of the Squeeze-and-Excitation (SE) attention mechanism optimizes feature weighting, allowing the model to focus more accurately on key regions of the image, even in the presence of complex backgrounds. The proposed model demonstrates a significant performance improvement, achieving an average precision (mAP) of 92.77%, which marks a substantial enhancement over the original YOLOv8. Additionally, the model reduces computational cost by approximately one-third in terms of FLOPS, maintaining high accuracy while greatly decreasing the number of parameters, thus offering a more robust and resource-efficient solution. For tongue crack detection, the mAP increases to 91.34%, with notable improvements in F1 score, precision, and recall. Teeth mark detection also sees a significant boost, achieving an mAP of 94.21%. These advancements underscore the model’s outstanding performance in TCM tongue image analysis, providing a more accurate, efficient, and reliable tool for clinical diagnostic applications.

A Multi-Branch Convolution and Dynamic Weighting Method for Bearing Fault Diagnosis Based on Acoustic–Vibration Information Fusion

Rolling bearings, as critical components of rotating machinery, directly affect the reliability and efficiency of the system. Due to extended operation under high load, harsh environmental conditions, and continuous use, bearings become more susceptible to failure, leading to a higher likelihood of malfunction. To prevent sudden failures, reduce downtime, and optimize maintenance strategies, early and accurate diagnosis of rolling bearing faults is essential. Although existing methods have achieved certain success in processing acoustic and vibration signals, they still face challenges such as insufficient feature fusion, inflexible weight allocation, lack of effective feature selection mechanisms, and low computational efficiency. To address these challenges, we propose a dynamic weighted multimodal fault diagnosis model based on the fusion of acoustic and vibration information. This model aims to enhance feature fusion, dynamically adapt to signal characteristics, optimize feature selection, and reduce computational complexity. The model incorporates an adaptive fusion method based on a multi-branch convolutional structure, enabling unified processing of both acoustic and vibration signals. At the same time, a cross-modal dynamic weighted fusion mechanism is employed, allowing the real-time adjustment of weight distribution based on signal characteristics. By utilizing an attention mechanism for dynamic feature selection and weighting, the robustness of classification is further improved. Additionally, when processing acoustic signals, a depthwise separable convolutional network is used, effectively reducing computational complexity. Experimental results demonstrate that our method significantly outperforms other algorithms in terms of convergence speed and final performance. Additionally, the accuracy curve during training showed minimal fluctuation, reflecting higher robustness. The model achieved over 99% diagnostic accuracy under all signal-to-noise ratio (SNR) conditions, showcasing exceptional robustness and noise resistance in both noisy and high-SNR environments. Furthermore, its superiority across different data scales, especially in small-sample learning and stability, highlights its strong generalization capability.

Backbone extraction through statistical edge filtering: A comparative study

The backbone extraction process is pivotal in expediting analysis and enhancing visualization in network applications. This study systematically compares seven influential statistical hypothesis-testing backbone edge filtering methods (Disparity Filter (DF), Polya Urn Filter (PF), Marginal Likelihood Filter (MLF), Noise Corrected (NC), Enhanced Configuration Model Filter (ECM), Global Statistical Significance Filter (GloSS), and Locally Adaptive Network Sparsification Filter (LANS)) across diverse networks. A similarity analysis reveals that backbones extracted with the ECM and DF filters exhibit minimal overlap with backbones derived from their alternatives. Interestingly, ordering the other methods from GloSS to NC, PF, LANS, and MLF, we observe that each method’s output encapsulates the backbone of the previous one. Correlation analysis between edge features (weight, degree, betweenness) and the test significance level reveals that the DF and LANS filters favor high-weighted edges while ECM assigns them lower significance to edges with high degrees. Furthermore, the results suggest a limited influence of the edge betweenness on the filtering process. The backbones global properties analysis (edge fraction, node fraction, weight fraction, weight entropy, reachability, number of components, and transitivity) identifies three typical behavior types for each property. Notably, the LANS filter preserves all nodes and weight entropy. In contrast, DF, PF, ECM, and GloSS significantly reduce network size. The MLF, NC, and ECM filters preserve network connectivity and weight entropy. Distribution analysis highlights the PU filter’s ability to capture the original weight distribution. NC filter closely exhibits a similar capability. NC and MLF filters excel for degree distribution. These insights offer valuable guidance for selecting appropriate backbone extraction methods based on specific properties.

Automatic feature selection and weighting in molecular systems using Differentiable Information Imbalance

Feature selection is essential in the analysis of molecular systems and many other fields, but several uncertainties remain: What is the optimal number of features for a simplified, interpretable model that retains essential information? How should features with different units be aligned, and how should their relative importance be weighted? Here, we introduce the Differentiable Information Imbalance (DII), an automated method to rank information content between sets of features. Using distances in a ground truth feature space, DII identifies a low-dimensional subset of features that best preserves these relationships. Each feature is scaled by a weight, which is optimized by minimizing the DII through gradient descent. This allows simultaneously performing unit alignment and relative importance scaling, while preserving interpretability. DII can also produce sparse solutions and determine the optimal size of the reduced feature space. We demonstrate the usefulness of this approach on two benchmark molecular problems: (1) identifying collective variables that describe conformations of a biomolecule, and (2) selecting features for training a machine-learning force field. These results show the potential of DII in addressing feature selection challenges and optimizing dimensionality in various applications. The method is available in the Python library DADApy.

Construction of a prediction and visualization system for cognitive impairment in elderly COPD patients based on self-assigning feature weights and residual evolution model.

Assessing cognitive function in patients with chronic obstructive pulmonary disease (COPD) is crucial for ensuring treatment efficacy and avoiding moderate cognitive impairment (MCI) or dementia. We aimed to build better machine learning models and provide useful tools to provide better guidance and assistance for COPD patients' treatment and care. A total of 863 COPD patients from a local general hospital were collected and screened, and they were separated into two groups: cognitive impairment (356 patients) and cognitively normal (507 patients). The Montreal Cognitive Assessment (MoCA) was used to test cognitive function. The swarm intelligence optimization algorithm (SIOA) was used to direct feature weighting and hyperparameter optimization, which were considered simultaneous activities. A self-assigning feature weights and residual evolution (SAFWRE) algorithm was built on the concept of linear and nonlinear information fusion. The best method in SIOA was the circle search algorithm. On the training set, SAFWRE's ROC-AUC was 0.9727, and its PR-AUC was 0.9663; on the test set, SAFWRE's receiver operating characteristic-area under curve (ROC-AUC) was 0.9243, and its precision recall-area under curve (PR-AUC) was 0.9059, and its performance was much superior than that of the control technique. In terms of external data, the classification and prediction performance of various models are comprehensively evaluated. SAFWRE has the most excellent classification performance, with ROC-AUC of 0.8865 and pr-auc of 0.8299. This work develops a practical visualization system based on these weight attributes which has strong application importance and promotion value.

A Dissimilarity Measure Powered Feature Weighted Fuzzy C-Means Algorithm for Gene Expression Data

A Dissimilarity Measure Powered Feature Weighted Fuzzy C-Means Algorithm for Gene Expression Data