计算机科学最新文献

{"title":"Global and local co-attention networks enhanced by learning state for knowledge tracing","authors":"Xinhua Wang,&nbsp;Yibang Cao,&nbsp;Liancheng Xu,&nbsp;Ke Sun","doi":"10.1007/s10489-025-06463-9","DOIUrl":"10.1007/s10489-025-06463-9","url":null,"abstract":"<div><p>In intelligent tutoring systems, knowledge tracing (KT) stands as a pivotal technology for facilitating personalized learning among students. Effectively capturing the continually evolving knowledge mastery states of students poses a formidable challenge in KT prediction. Traditional KT methods typically model students’ global knowledge mastery states solely based on the chronological sequence of their historical interactions, neglecting the significance of their current learning state and the inherent interplay between global and local knowledge mastery states. To bridge these gaps, this paper introduces a novel Learning State Enhanced Co-attention Model (LSEKT) for knowledge tracing. In terms of methodology, we contend that a student’s recent answering behavior is intricately tied to implicit learning states. Consequently, we devise a learning state extraction network to capture the student’s current learning state. Furthermore, to construct a more robust and interdependent representation of both global and local knowledge mastery states, we integrate a co-attention network. This network enhances the attention paid to pertinent knowledge points across both global and local scales, thereby adeptly capturing the underlying connections between global and local interaction sequences. Concurrently, we incorporate contrastive learning as an auxiliary task within our model to bolster its predictive prowess. Ultimately, we evaluated our approach through extensive experiments on four widely used datasets. The experimental outcomes underscore the remarkable performance of our model across diverse evaluation metrics, emphasizing the effectiveness of our proposed LSEKT model.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 7","pages":""},"PeriodicalIF":3.4,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Spin Recovery of High-Angle-of-Attack Aircraft With Altitude Gain Reduction in the Presence of Aerodynamic Uncertainty: A MIMO Super-Twisting Sliding Mode Approach","authors":"Ahmad Bagheri,&nbsp;Mohammad Danesh","doi":"10.1049/cth2.70018","DOIUrl":"https://doi.org/10.1049/cth2.70018","url":null,"abstract":"<p>To recover steady, straight-level flight of a high-angle-of-attack aircraft from its oscillatory spin, a MIMO super-twisting sliding control approach is proposed in this study. Since at high angles of attack, the aerodynamics governing the aircraft is highly nonlinear, tabulated data are utilised to ensure the validity of the results up to an angle of attack of 90°. Regarding uncertain aerodynamic coefficients, the robustness of the control approach is necessary. It is shown that the first-order classical sliding control and power rate reaching law methods are successful approaches to recover an aircraft from its state of spin in the absence of aerodynamic parameter uncertainties. However, in the presence of these uncertainties, chattering affects their performance and the altitude required to perform the recovery manoeuvre, referred to as altitude gain, significantly increases. To overcome these issues, a second-order sliding control algorithm is proposed in this study. The system outputs are considered as roll, pitch, rate of yaw change to attain level flight, and rate of change of altitude to assure straight flight. Thus, a 4 × 4 super-twisting SMC scheme is developed. Finite-time convergence of sliding variables, which guarantees asymptotic stability of the aircraft control system, is proven via the Lyapunov direct method. Simulation results illustrate that the proposed control algorithm serves not only as a reliable approach to perform the recovery manoeuvre but also as a highly effective method to overcome aerodynamic uncertainties without inducing chattering in control inputs. In addition, it enables the recovery manoeuvre to be performed with lower altitude gain.</p>","PeriodicalId":50382,"journal":{"name":"IET Control Theory and Applications","volume":"19 1","pages":""},"PeriodicalIF":2.2,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1049/cth2.70018","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143770404","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"StealthMask: Highly stealthy adversarial attack on face recognition system","authors":"Jian-Xun Mi,&nbsp;Mingxuan Chen,&nbsp;Tao Chen,&nbsp;Xiao Cheng","doi":"10.1007/s10489-025-06511-4","DOIUrl":"10.1007/s10489-025-06511-4","url":null,"abstract":"<div><p>Convolutional Neural Networks (CNNs) based on deep learning algorithms are widely used in real-world scenarios. However, these networks are vulnerable to adversarial examples-maliciously crafted inputs that can cause the model to make incorrect predictions. The existence of adversarial examples presents a significant challenge to the field of deep learning, with profound implications for various aspects of our lives. In face recognition technology, adversarial examples pose a substantial security risk. In this paper, we propose a novel method for generating adversarial patches designed to be worn as masks. The perturbed mask is crafted to deceive face recognition models, thereby highlighting the security vulnerabilities inherent in this technology. Our experimental results demonstrate that the mask generated by the proposed method effectively misleads the face recognition system, achieving high attack success rates while maintaining necessary stealthiness and transferability. Moreover, our method successfully attacks commercial face recognition systems and real-world access control systems, exposing the vulnerabilities of existing face recognition technologies in security-critical applications. Notably, compared to traditional methods, our proposed method emphasizes the stealthiness of the adversarial mask more than traditional methods. To account for physical-world factors, such as distortion, rotation, and deformations, we integrate a specifically designed loss function, thereby enhancing the method’s stability and reliability in practical scenarios.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 7","pages":""},"PeriodicalIF":3.4,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143769832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"End-to-end privacy-preserving image retrieval in cloud computing via anti-perturbation attentive token-aware vision transformer","authors":"Qihua Feng ,&nbsp;Zhixun Lu ,&nbsp;Chaozhuo Li ,&nbsp;Feiran Huang ,&nbsp;Jian Weng ,&nbsp;Philip S. Yu","doi":"10.1016/j.inffus.2025.103153","DOIUrl":"10.1016/j.inffus.2025.103153","url":null,"abstract":"<div><div>Privacy-Preserving Image Retrieval (PPIR) has gained popularity among users who upload encrypted personal images to remote servers, enabling image retrieval anytime and anywhere with privacy protection. Existing PPIR suggests extracting features from cipher-images through artificially-designed methods or Convolutional Neural Networks (CNNs). Nonetheless, manual feature engineering entails additional human effort, while CNNs are sensitive to spatial permutations as they primarily manipulate local texture features. To this end, we propose an innovative end-to-end PPIR, which not only eliminates the hassle of manual features but also enables learning expressive cipher-image representations. Specifically, since Vision Transformer (ViT) exhibits excellent robustness against permutation and occlusion in images, we elaborately design an Attentive Token-Aware (ATA) ViT model and hierarchical image block encryptions, which organically complement each other in an end-to-end system. The ATA module effectively learns informative block tokens and pays less attention to trivial and noisy encrypted blocks. Besides, to deal with the problem that the generalization of the model could be hindered by data desert, we adaptively construct the cipher-image augmentations by random block swapping and block erasing, aligning with our encryption operation. Extensive experiments on two datasets validate the superior retrieval accuracy and competitive image privacy protection performance of our proposed scheme.</div></div>","PeriodicalId":50367,"journal":{"name":"Information Fusion","volume":"121 ","pages":"Article 103153"},"PeriodicalIF":14.7,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143768934","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Three-level-fused attribute reductions based on three-view uncertainty measures of three-view weighted neighborhood rough sets","authors":"Jiang Chen ,&nbsp;Xianyong Zhang ,&nbsp;Xiao Tang ,&nbsp;Xiaoling Yang ,&nbsp;Zhiying Lv","doi":"10.1016/j.engappai.2025.110644","DOIUrl":"10.1016/j.engappai.2025.110644","url":null,"abstract":"<div><div>Attribute reductions facilitate classification learning, and they rely on uncertainty measures related to knowledge granulations of various rough sets. Weighted neighborhood rough sets (WNRSs) introduce attribute weights to optimize granular structures and improve neighborhood models, but their current attribute reductions and corresponding algorithms utilize algebra weights and dependency degrees from only a single algebraic perspective. For enriching WNRSs and their attribute reductions, three-view attribute weights and three-view granulation measures are constructed from algebraic, informational, and algebra-information-fused viewpoints, and thus <span><math><mrow><mn>3</mn><mo>×</mo><mn>3</mn><mo>=</mo><mn>9</mn></mrow></math></span> attribute reductions are systematically proposed to generate heuristic algorithms on three-level fusion. First based on existing algebraic weights, informational weights are supplemented by information entropy, while fused weights are constructed by algebra-information integration; thus, three-view weights induce three-view weighted neighborhood granulations and three-view WNRSs, and the latter’s granulation monotonicity is achieved. Then based on WNRSs and dependency degrees, decision entropies are supplemented by information function on roughness, while fused measures are constructed by multiplication operation; thus, three-view measures are formulated to acquire granulation monotonicity. Furthermore, three-view weights and measures motivate attribute reductions of WNRSs, and <span><math><mrow><mn>3</mn><mo>×</mo><mn>3</mn><mo>=</mo><mn>9</mn></mrow></math></span> heuristic reduction algorithms are systematically designed, thus extending and improving the existing algorithm; in terms of fusion strength, these 9 algorithms exhibit a three-level structure of <span><math><mrow><mn>4</mn><mo>+</mo><mn>4</mn><mo>+</mo><mn>1</mn></mrow></math></span> on non-fusion, single-fusion, double-fusion. Finally through data experiments, relevant uncertainty measures and attribute reductions are validated; all 9 reduction algorithms are comprehensively compared in classification learning, the new algorithms generally outperform the contrastive algorithm, and the double-fused algorithm on fused weights and measures acquires the best performance.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"151 ","pages":"Article 110644"},"PeriodicalIF":7.5,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Topology-preserving and structure-aware (hyper)graph contrastive learning for node classification","authors":"Minhao Zou,&nbsp;Zhongxue Gan,&nbsp;Yutong Wang,&nbsp;Junheng Zhang,&nbsp;Chun Guan,&nbsp;Siyang Leng","doi":"10.1007/s10489-025-06491-5","DOIUrl":"10.1007/s10489-025-06491-5","url":null,"abstract":"<div><p>Recently, graph contrastive learning (GCL) has attracted considerable attention, establishing a new paradigm for learning graph representations in the absence of human annotations. While notable advancements have been made, simultaneous consideration of both graphs and hypergraphs remains rare. This limitation arises because graphs and hypergraphs encode connectivity differently, making it challenging to develop a unified structure augmentation strategy. Conventional structure augmentation methods like adding or removing edges risk imperiling intrinsic topological traits and introducing adverse distortions such as disconnected subgraphs or isolated nodes. In this work, we propose a framework of contrastive learning on graphs and hypergraphs, named as UniGCL, to address these challenges by leveraging a unified adjacency representation that enables simultaneous modeling of pairwise and higher-order relationships. In particular, two structure augmentation methods are developed to perturb graph structure weights instead of altering connectivity, thereby preserving both graph and hypergraph topology while generating diverse augmented views. Furthermore, a structure-aware contrastive loss is proposed, which incorporates gradient perturbation techniques to enhance the model’s ability to capture fine-grained structural dependencies in (hyper)graphs. Extensive experiments are conducted on six real-world graph datasets and nine representative hypergraph datasets to evaluate the performance of the proposed framework. The results demonstrate that UniGCL achieves superior node classification performance compared to the advanced graph and hypergraph contrastive learning methods, across datasets with different homophilic extents and limited annotations. Additionally, ablation studies validate the effectiveness of our structure-preserving augmentations and structure-aware contrastive loss in enhancing performance.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 7","pages":""},"PeriodicalIF":3.4,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Unsupervised learning with physics informed graph networks for partial differential equations","authors":"Lin Lu,&nbsp;Yiye Zou,&nbsp;Jingyu Wang,&nbsp;Shufan Zou,&nbsp;Laiping Zhang,&nbsp;Xiaogang Deng","doi":"10.1007/s10489-025-06479-1","DOIUrl":"10.1007/s10489-025-06479-1","url":null,"abstract":"<div><p>Natural physical phenomena are commonly expressed using partial differential equations (PDEs), in domains such as fluid dynamics, electromagnetism, and atmospheric science. These equations typically require numerical solutions under given boundary conditions. There is a burgeoning interest in the exploration of neural network methodologies for solving PDEs, mainly based on automatic differentiation methods to learn the PDE-solving process, which means that the model needs to be retrained when the boundary conditions of PDE are changed. However, automatic differentiation requires substantial memory resources to facilitate the training regimen. Moreover, a learning objective that is tailored to the solution process often lacks the flexibility to extend to boundary conditions; thereby limiting the solution’s overall precision. The method proposed in this paper introduces a graph neural network approach, embedded with physical information, mainly for solving Poisson’s equation. An approach is introduced that reduces memory usage and enhances training efficiency through an unsupervised learning methodology based on numerical differentiation. Concurrently, by integrating boundary conditions directly into the neural network as supplementary physical information, this approach ensures that a singular model is capable of solving PDEs across a variety of boundary conditions. To address the challenges posed by more complex network inputs, the introduction of graph residual connections serves as a strategic measure to prevent network overfitting and to elevate the accuracy of the solutions provided. Experimental findings reveal that, despite having 30 times more training parameters than the Physics-Informed Neural Networks (PINN) model, the proposed model consumes 2.2% less memory than PINN. Additionally, generalization in boundary conditions has been achieved to a certain extent. This enables the model to solve partial differential equations with different boundary conditions, a capability that PINN currently lacks. To validate the solving capability of the proposed method, it has been applied to the model equation, the Sod shock tube problem, and the two-dimensional inviscid airfoil problem. In terms of the solution accuracy of the model equations, the proposed method outperforms PINN by 30% to four orders of magnitude. Compared to the traditional numerical method, the Finite Element Method (FEM), the proposed method also shows an order of magnitude improvement. Additionally, when compared to the improved version of PINN, TSONN, our method demonstrates certain advantages. The forward problem of the Sod shock tube, which PINN is currently unable to solve, is successfully handled by the proposed method. For the airfoil problem, the results are comparable to those of PINN.</p></div>","PeriodicalId":8041,"journal":{"name":"Applied Intelligence","volume":"55 7","pages":""},"PeriodicalIF":3.4,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143761711","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Two-sided matching optimization model for green housing technology selection based on hesitant 2-tuple linguistic rough numbers","authors":"Musavarah Sarwar ,&nbsp;Ghous Ali ,&nbsp;Wajeeha Gulzar ,&nbsp;Muhammad Akram ,&nbsp;Dragan Pamucar","doi":"10.1016/j.engappai.2025.110559","DOIUrl":"10.1016/j.engappai.2025.110559","url":null,"abstract":"<div><div>In two-sided matching decision problems, the matching objects with different knowledge, experiences, and cultures provide linguistic assessments using diverse or multi-granular sets with a factor that the information provided is hesitant in nature due to different opinions given by experts. In the proposed approach, the hesitant 2-tuple linguistic information is integrated with rough approximations to develop the two novel approaches called hesitant rough numbers and hesitant 2-tuple linguistic rough numbers. The proposed novel approximations are implemented on a two-sided matching optimization model to study hesitant multi-granular uncertainty. Firstly, the matching objects provide their evaluations in the form of hesitant multi-granular terms converted into hesitant 2-tuple linguistic rough numbers. Secondly, certain optimization models based on hesitant 2-tuple linguistic rough approximations are constructed to compute the criteria weights using incomplete information. In hesitant 2-tuple linguistic rough optimization models, the maximizing deviation technique is used to find the distance between two proposed novel coefficients. To maximize the level of satisfaction with matching objects, a hesitant 2-tuple linguistic rough optimization model is developed to evaluate the overall satisfaction degree and stability of matching objects. The significance of the proposed two-sided matching optimization model is illustrated with a case study of matching between the green building technology supply and demand. The out-performance of the proposed model is highlighted by a comparison analysis with existing approaches to analyze that it can provide hesitant multi-granular rough flexibility and deals with incomplete information regarding criterion weights.</div></div>","PeriodicalId":50523,"journal":{"name":"Engineering Applications of Artificial Intelligence","volume":"151 ","pages":"Article 110559"},"PeriodicalIF":7.5,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760228","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Advancing recommendation systems with DeepMF and hybrid sentiment analysis: Deep learning and Lexicon-based integration","authors":"Nossayba Darraz ,&nbsp;Ikram Karabila ,&nbsp;Anas El-Ansari ,&nbsp;Nabil Alami ,&nbsp;Mostafa El Mallahi","doi":"10.1016/j.eswa.2025.127432","DOIUrl":"10.1016/j.eswa.2025.127432","url":null,"abstract":"<div><div>In the hotel industry, ensuring customer satisfaction and providing personalized recommendations are crucial elements for creating a remarkable guest experience. However, traditional recommendation systems encounter several challenges that hinder their effectiveness. These challenges include cold start problems, where it is difficult to make recommendations for new or less-rated items, as well as data sparsity, which limits the availability of relevant information. Additionally, accurately interpreting the diverse sentiments expressed by customers in their reviews poses another significant challenge. This study tackles these challenges by integrating sentiment analysis into hotel recommendation systems, aiming to capture and analyze guest opinions and sentiments from their reviews. This study aims to enhance recommendation systems by integrating a hybrid sentiment analysis approach. The approach combines lexicon-based techniques and deep learning methodologies, using TextBlob with Bag of Words and a Multilayer Perceptron (MLP) algorithm to analyze the sentiment of textual data. The hybrid sentiment analysis approach exhibits an impressive accuracy rate of 88.63%, demonstrating its effectiveness in capturing sentiment from customer reviews. This integration enables recommendation systems to better understand and incorporate customer sentiments, leading to improved personalized recommendations. Moreover, we combine this hybrid sentiment analysis with DeepMF for collaborative hotel recommendations, which yields a remarkable Root Mean Square Error (RMSE) of 0.1. By integrating sentiment analysis into the recommendation system, we gain valuable insights into customer preferences, leading to improved recommendation quality and personalization. This research highlights the potential of sentiment analysis in optimizing customer experience management within the hotel industry, providing a valuable tool for enhancing guest satisfaction and engagement.</div></div>","PeriodicalId":50461,"journal":{"name":"Expert Systems with Applications","volume":"279 ","pages":"Article 127432"},"PeriodicalIF":7.5,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ensemble learning unlocking point load forecasting accuracy: A novel framework based on two-stage data preprocessing and improved multi-objective optimisation strategy","authors":"Jingmin Luan ,&nbsp;Qiyang Li ,&nbsp;Yuyan Qiu ,&nbsp;Weican Liu","doi":"10.1016/j.compeleceng.2025.110282","DOIUrl":"10.1016/j.compeleceng.2025.110282","url":null,"abstract":"<div><div>Accurate point power forecasts are critical for maintaining the security and stability of the grid. However, Load data is volatile and difficult to predict with high accuracy. To improve the stability and accuracy of the model, we proposed a novel hybrid load forecasting framework consisting of three modules. Firstly, we used <em>Successive Variational Mode Decomposition (SVMD)</em> in the data preprocessing module to extract trend features and denoise the data. To reduce the negative impact of feature redundancy on model predictions, we employed feature selection to identify four essential features to aid model training. Secondly, in the ensemble learning module, we address the limitations of single predictive models by combining the models <em>Back Propagation (BP)</em>, <em>Temporal Convolutional Network (TCN)</em>, <em>Bidirectional Long Short-Term Memory (BiLSTM)</em>, <em>Bidirectional Gated Recurrent Unit (BiGRU)</em> and <em>Transformer</em>. Finally, in the improved multi-objective optimisation algorithm module, we implemented various strategies to enhance the optimisation algorithm. We designed some experiments using three load datasets from Australia. The results demonstrated that the mean absolute percentage error values are below <strong><em>1.25%</em></strong>, with the best value reaching <strong><em>0.9256%</em></strong>. In contrast, the best result for the mean absolute percentage error of the baseline models was <strong><em>1.4431%</em></strong> in New South Wales. This represents a <strong><em>35.9%</em></strong> improvement in load forecasting performance with our proposed model, highlighting its superior accuracy compared to competing approaches. It shows that our forecasting framework is far better than other rivals.</div></div>","PeriodicalId":50630,"journal":{"name":"Computers & Electrical Engineering","volume":"124 ","pages":"Article 110282"},"PeriodicalIF":4.0,"publicationDate":"2025-04-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143760338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}