Algorithms

21 pages, 496 KB

Open AccessArticle

The AI Overtrust Paradox in Vulnerability Search: When Greater Trust in AI Reduces Perceived Search Effectiveness

by Boris Kriheli and Eugene Levner

Algorithms 2026, 19(9), 792; https://doi.org/10.3390/a19090792 - 15 Sep 2026

We investigate human overreliance on artificial intelligence (AI), a phenomenon in which users fail to critically verify AI outputs despite knowing the system is imperfect. To analyze this interaction, we model a cybersecurity scenario in which a human user and an AI chatbot [...] Read more.

We investigate human overreliance on artificial intelligence (AI), a phenomenon in which users fail to critically verify AI outputs despite knowing the system is imperfect. To analyze this interaction, we model a cybersecurity scenario in which a human user and an AI chatbot collaborate to locate a vulnerable node in a large-scale distributed network. We consider a search process in which the probability distribution of the node’s location is updated using Bayes’ rule. In each round, the chatbot sequentially scans the nodes and suggests a potentially vulnerable node; the user then decides whether to accept the detection and end the search or refine the query and continue searching. The goal is to find an efficient scanning algorithm that maximizes the user’s expected satisfaction with the search results. Finally, we provide a formal proof of the “more for less” paradox, demonstrating that reducing trust in AI can, under certain conditions, increase perceived search effectiveness. Full article

(This article belongs to the Special Issue Algorithms for Cyber Defense: From Cryptography to Behavioral Analysis)

41 pages, 15921 KB

Open AccessArticle

A Mobility Priority Index for Data-Driven National Bridge Preservation Prioritization

by Raj Bridgelall

Algorithms 2026, 19(9), 791; https://doi.org/10.3390/a19090791 - 15 Sep 2026

Abstract

Bridge and pavement management systems typically evaluate structural condition and roadway performance separately, limiting network-level consideration of infrastructure condition and traffic exposure. This study develops the Mobility Priority Index (MPI), a relative composite screening measure that integrates pavement roughness, bridge deck condition, current [...] Read more.

Bridge and pavement management systems typically evaluate structural condition and roadway performance separately, limiting network-level consideration of infrastructure condition and traffic exposure. This study develops the Mobility Priority Index (MPI), a relative composite screening measure that integrates pavement roughness, bridge deck condition, current traffic exposure, and projected traffic growth. A transparent and transferable workflow spatially linked the Highway Performance Monitoring System and National Bridge Inventory and applied sequential quality screening to 79,595 open, unrestricted mainline National Highway System bridges in the contiguous United States. The MPI was descriptively characterized using distributional analysis; spatial patterns and specification sensitivity were examined using local Getis–Ord

G_{i}^{*}

hotspot analysis and eight alternative formulations. The baseline identified 20,621 bridges (25.9%) as nominally significant exploratory hotspots, concentrated mainly in metropolitan regions and high-demand corridors. Across the alternative specifications, bridge-rank correlations ranged from 0.807 to 0.965, whereas hotspot-set Jaccard similarities ranged from 0.446 to 0.801, indicating greater sensitivity in hotspot membership than in overall rankings. The MPI does not measure realized delay, congestion, reliability, user cost, or investment benefit. Instead, it provides a nationally scalable screening method that identifies bridges and regions warranting further engineering and operational evaluation during preservation planning. Full article

(This article belongs to the Special Issue Transportation and Traffic Engineering)

► Show Figures

Figure 1

41 pages, 20383 KB

Open AccessArticle

A Hybrid Gabor–ViT Embedding with Gated Mamba Deep Learning Framework for Brain Tumor MRI Classification

by Baidaa Mutasher Rashed, Zahoor M. Aydam and Nidhal Khdhair El Abbadi

Algorithms 2026, 19(9), 790; https://doi.org/10.3390/a19090790 - 15 Sep 2026

Abstract

Brain tumor classification from MRI is an important task in medical image analysis. This study introduces a new hybrid deep learning method that integrates Gabor filtering for the improvement in texture feature extraction, the Vision Transformer (ViT) framework for extraction of the global [...] Read more.

Brain tumor classification from MRI is an important task in medical image analysis. This study introduces a new hybrid deep learning method that integrates Gabor filtering for the improvement in texture feature extraction, the Vision Transformer (ViT) framework for extraction of the global context dependency and fine-grained spatial features, and a novel Gated Mamba to enhance feature classification efficiency. The framework was first evaluated in-distribution on the Kaggle dataset, which contains 5712 slices of brain MRI images from four different classes using accuracy, precision, recall, F1-score, AUC, Std, and 95% confidence intervals (CIs), with a mean image-level accuracy of 99.62%. Because the Kaggle dataset is known to be near-saturated and no patient identifiers are available, this in-distribution result is treated as a sanity check rather than a measure of clinical generalization; pHash deduplication was used to remove visually identical images, but this does not provide patient-level grouping, and residual patient-level leakage cannot be fully excluded. The central evaluation of this study is therefore a cross-dataset validation on the “figshare brain mri” dataset, which uses a different scanner and acquisition protocol and was never used during training and model selection; the proposed model on this dataset achieves 87.23% accuracy with reasonable confidence intervals, a decline of 12.39 percentage points relative to the in-distribution result that reflects the domain-shift gap typical of this task. The proposed hybrid approach performs competitively against established methods in-distribution, while the cross-dataset results highlight generalization challenges that motivate the failure analysis and limitations discussed in this paper. Full article

(This article belongs to the Special Issue Algorithms for Computer Aided Diagnosis: 3rd Edition)

► Show Figures

Figure 1

14 pages, 588 KB

Open AccessCommunication

Optimizing Infrastructure Inspection Spans Under Assessor Bias: A Stochastic Simulation Algorithm for Mitigating Fat-Tail Social Risks

by Yasuko Kawahata, Durga Chavali, Noriaki Maeda and Shunsuke Hatadani

Algorithms 2026, 19(9), 789; https://doi.org/10.3390/a19090789 - 15 Sep 2026

Abstract

The rapid aging of social infrastructure constructed during Japan’s high-growth period presents a critical national challenge. Following a major tunnel collapse in 2012, Japan mandated a uniform “once-every-5-years” visual inspection policy. While establishing a safety baseline, this static span, largely an administrative compromise, [...] Read more.

The rapid aging of social infrastructure constructed during Japan’s high-growth period presents a critical national challenge. Following a major tunnel collapse in 2012, Japan mandated a uniform “once-every-5-years” visual inspection policy. While establishing a safety baseline, this static span, largely an administrative compromise, fails to address the severe non-linear degradation dynamics exacerbated by extreme environmental shocks in disaster-prone regions. This study challenges the adequacy of this uniform interval by developing a probabilistic framework that integrates non-linear degradation dynamics with Multiple Dependent Competing Failure Processes (MDCFPs). To bridge the gap between engineering predictions and real-world operations, we incorporate assessor bias through Bayesian modeling of imperfect inspections. Furthermore, we transition from conventional agency-cost minimization to Social Life Cycle Cost (LCC) optimization, explicitly quantifying public health risks and socio-economic externalities. Large-scale Monte Carlo simulations (

N = 10^{7}

) reveal a critical divergence: while conventional agency-cost optimization misleadingly suggests an optimal preventive intervention point of 5–6 years, our rigorous evaluation identifies the absolute physical safety limit at exactly 2.79 years. This demonstrates that the practical 5-year default already severely breaches physical thresholds, exposing society to catastrophic fat-tail risks. These findings substantiate the urgent shift toward condition-based dynamic scheduling and highlight the necessity of implementing policy safeguards to ensure social equity for vulnerable communities. Full article

(This article belongs to the Section Algorithms for Multidisciplinary Applications)

► Show Figures

Figure 1

22 pages, 5159 KB

Open AccessArticle

Interpretable Heart Disease Prediction: Optimizing Machine Learning Models via Metaheuristic Ivy Algorithm

by Yang Jiang, Zihao Zuo, Rui Liang, Jiabin Xu, Hong Jiang, Zhigang Ding, Yanhong Peng and Cong Li

Algorithms 2026, 19(9), 788; https://doi.org/10.3390/a19090788 - 14 Sep 2026

Abstract

Cardiovascular diseases remain a major global health burden, making the development of accurate and interpretable prediction models important for clinical decision support. In this study, the metaheuristic Ivy Algorithm was employed to perform two-stage hyperparameter optimization for five machine learning classifiers, namely ID3, [...] Read more.

Cardiovascular diseases remain a major global health burden, making the development of accurate and interpretable prediction models important for clinical decision support. In this study, the metaheuristic Ivy Algorithm was employed to perform two-stage hyperparameter optimization for five machine learning classifiers, namely ID3, SVM, RF, XGBoost, and LightGBM. The proposed framework was evaluated on the publicly available Cleveland and Statlog heart disease datasets using outer stratified 10-fold cross-validation. The results showed that IVYA-based optimization improved the predictive performance of all five classifiers to varying degrees. Among them, IVYA-LightGBM achieved the best overall performance, with mean AUC, Accuracy, Precision, Recall, and F1-score values of 0.945, 0.907, 0.931, 0.864, and 0.893, respectively. Paired Wilcoxon signed-rank tests based on the fold-wise results indicated that the improvements in AUC were statistically significant in most model–dataset comparisons. In addition, under consistent experimental settings, IVYA was compared with five widely used metaheuristic optimization algorithms and achieved the highest AUC, Recall, and F1-score, while requiring the shortest average runtime. To enhance model interpretability, SHAP analysis was further incorporated to quantify the contributions of different clinical features to the model predictions and improve the transparency of the prediction process. Overall, IVYA-LightGBM achieved a favorable balance among predictive performance, computational efficiency, and interpretability. Nevertheless, further validation on larger and more diverse clinical datasets is required before practical clinical application. Full article

(This article belongs to the Special Issue Computational Intelligence and Nature Inspired Algorithms)

► Show Figures

Figure 1

16 pages, 1733 KB

Open AccessArticle

A Machine Learning Framework for EEG-Based Epileptic Seizure Classification Using Time–Frequency Feature Analysis

by Faiza Dad, Fred Lacy and Yasser Ismail

Algorithms 2026, 19(9), 787; https://doi.org/10.3390/a19090787 - 14 Sep 2026

Abstract

Epilepsy affects approximately 50 million people worldwide, and timely, accurate seizure detection is essential for clinical management. Electroencephalogram (EEG) signals carry rich temporal and spectral information about brain dynamics, yet manual interpretation is time-consuming and expert-dependent. This paper presents a complete automated EEG [...] Read more.

Epilepsy affects approximately 50 million people worldwide, and timely, accurate seizure detection is essential for clinical management. Electroencephalogram (EEG) signals carry rich temporal and spectral information about brain dynamics, yet manual interpretation is time-consuming and expert-dependent. This paper presents a complete automated EEG classification pipeline combining signal processing feature engineering with supervised machine learning to distinguish epileptic from non-epileptic brain states. Using the UCI Epileptic Seizure Recognition Dataset (11,500 segments of approximately 1.025 s and five classes at 173.61 Hz), we extract 31 features spanning the time domain (mean, standard deviation, RMS, peak-to-peak, zero-crossing rate, skewness, kurtosis, Hjorth activity, mobility, complexity, line length, mean absolute deviation, and derivative variance) and the frequency domain (DFT magnitude statistics and Welch PSD statistics, spectral entropy, spectral centroid, spectral bandwidth, total power, EEG band powers (delta through gamma), and inter-band power ratios). Four classifiers—logistic regression (LR), random forest (RF), support vector machine (SVM), and gradient boosting (GB)—are trained and evaluated on a binary task (seizure versus non-seizure) and a five-class task. For binary classification, the SVM achieves the highest accuracy of 99.13% with an F1 score of 0.978. For multiclass classification, gradient boosting achieves 81.35% accuracy. Feature importance analysis identifies the Hjorth complexity, spectral centroid, and theta-band power as the most discriminative features. A comprehensive ablation study further confirms that the combined 31-feature representation outperforms either domain in isolation by up to 4.2 percentage points on the multiclass task. Unlike less interpretable deep learning models, the proposed signal processing pipeline is computationally efficient and provides transparent feature-based classification on this benchmark dataset. The complete pipeline is implemented in Python 3.13.15 and is fully reproducible. Full article

(This article belongs to the Special Issue Machine Learning Techniques for Brain Data Analysis Using EEG, EMG or Image Data)

► Show Figures

Figure 1

29 pages, 2749 KB

Open AccessArticle

A Hybrid PSO–GA Algorithm for Adaptive Decentralized Sequencer Selection and DAG-Aware Workload Coordination in Layer-2 Blockchain Networks

by Unnikrishnan Kalleli Narayanan and P. Victer Paul

Algorithms 2026, 19(9), 786; https://doi.org/10.3390/a19090786 - 14 Sep 2026

Abstract

Decentralized Layer-2 sequencing requires fixed-size committees to balance performance, fairness, and resilience while downstream transaction tasks remain dependency-constrained. This study formulates committee selection as a bounded seven-objective discrete optimization problem and combines a feasibility-preserving hybrid particle swarm optimization–genetic algorithm (PSO–GA) search with directed [...] Read more.

Decentralized Layer-2 sequencing requires fixed-size committees to balance performance, fairness, and resilience while downstream transaction tasks remain dependency-constrained. This study formulates committee selection as a bounded seven-objective discrete optimization problem and combines a feasibility-preserving hybrid particle swarm optimization–genetic algorithm (PSO–GA) search with directed acyclic graph scheduling and post-selection Byzantine fault tolerance (BFT) audits. Across 72 matched model-based runs, hybrid PSO–GA produced a 16.1% higher score, 21.9% lower latency, 47.0% higher throughput, 9.6% lower energy, and 6.2% lower effective gas than the controlled stake-only BFT baseline. It also exceeded random eligible selection, PSO-only, MOPSO, and NSGA-II; however, GA-only attained the highest scalar score (0.8299 versus 0.8266). The hybrid is therefore interpreted as a competitive feasibility-preserving trade-off, not a universally superior optimizer. Under high dependency and task heterogeneity, the scheduler reduced mean makespan to 32.75 from 34.38 for FIFO-ready and 49.59 for round-robin scheduling. Targeted attack simulations could exceed the BFT budget. The results support the framework as a reproducible decision model, while testbed deployment, authenticated telemetry, and incentive mechanisms remain necessary for protocol-level validation. Full article

► Show Figures

Graphical abstract

41 pages, 6308 KB

Open AccessArticle

Performance Comparison Among Classical Metaheuristic Algorithms for Stochastic Last-Mile Delivery Routing Problem

by Bonginkosi A. Thango and Osayuwamen Omoruyi

Algorithms 2026, 19(9), 785; https://doi.org/10.3390/a19090785 - 13 Sep 2026

Abstract

Last-mile delivery routing requires simultaneous control of distance, travel time, service deadlines, vehicle capacity, workload balance, operating cost, emissions, and uncertainty. This study provides a controlled low-budget comparison of ten classical metaheuristics for a LaDe-calibrated stochastic CVRPTW. The GA, DE, PSO, ACO, ABC, [...] Read more.

Last-mile delivery routing requires simultaneous control of distance, travel time, service deadlines, vehicle capacity, workload balance, operating cost, emissions, and uncertainty. This study provides a controlled low-budget comparison of ten classical metaheuristics for a LaDe-calibrated stochastic CVRPTW. The GA, DE, PSO, ACO, ABC, SA, GWO, WOA, TLBO, and JAYA used the same random-key representation, capacity-aware split decoder, repair rules, CRN scenarios, and exactly 50 objective evaluations per run. The experiment comprised 1065 fixed design conditions, 30 seeded runs per algorithm-condition cell, and 319,500 metaheuristic runs, with NNS, Clarke–Wright Savings, and OR-Tools Guided Local Search as benchmarks. PSO achieved the lowest grand-mean weighted logistics cost, whereas the WOA achieved the lowest median and best Friedman mean rank. The grand-mean difference between the WOA and PSO was only 125.2 objective units, and the context winner changed across the three cities and five customer-size levels (PSO: 4, WOA: 2, and JAYA: 2). Accordingly, corrected p-values are treated as conditional design diagnostics, while practical interpretation prioritises paired magnitude, stratified consistency, feasibility, and benchmark proximity. A design-stratified sensitivity analysis showed that the preferred method changed across cities and customer sizes. Crucially, every metaheuristic and benchmark recorded 0% full feasibility under the common evaluator; capacity- and lateness-violation profiles are therefore elevated to primary outcomes, and no algorithm is claimed to be operationally superior. The 50-evaluation protocol is interpreted as an early-budget screening regime rather than evidence of asymptotic convergence. The results show that algorithm choice is conditional on the tested data, weights, problem sizes, uncertainty settings, and evaluation budget and that feasibility-preserving decoding is more important for deployment than small differences in penalised objective value. Full article

► Show Figures

Figure 1

34 pages, 633 KB

Open AccessArticle

Tensor-Train Formulation and Numerical Study of the Peaceman–Rachford and Douglas–Rachford Splitting Schemes for Multidimensional Parabolic Evolution Equations

by Gianmarco Manzini

Algorithms 2026, 19(9), 784; https://doi.org/10.3390/a19090784 - 11 Sep 2026

Abstract

We formulate the Peaceman–Rachford (PR) and Douglas–Rachford (DR) operator splitting schemes for multidimensional parabolic evolution equations in the tensor-train (TT) format. The resulting TT-PR and TT-DR algorithms maintain the evolving state and the split operators in TT form, solve the implicit subproblems approximately [...] Read more.

We formulate the Peaceman–Rachford (PR) and Douglas–Rachford (DR) operator splitting schemes for multidimensional parabolic evolution equations in the tensor-train (TT) format. The resulting TT-PR and TT-DR algorithms maintain the evolving state and the split operators in TT form, solve the implicit subproblems approximately with the Alternating Minimal Energy (AMEn) method, and control TT ranks through preventive rounding and a rank cap enforced within the inner solves. We analyze per-step computational complexity, practical rank behavior, and total error decomposition, and we validate the algorithms on benchmark problems in two to eight spatial dimensions, with additional higher-dimensional scaling illustrations. Full article

(This article belongs to the Section Algorithms for Multidisciplinary Applications)

► Show Figures

Figure 1

32 pages, 8453 KB

Open AccessArticle

Joint Location and Capacity Optimization of Electric Vehicle Charging and Battery-Swapping Stations Using Random Forest Surrogate-Assisted NSGA-III

by Zihan Li, Bo Yang, Huanming Zhang, Xiangyu Zhao, Yuanweiji Hu, Junyu Liang and Xuehao He

Algorithms 2026, 19(9), 783; https://doi.org/10.3390/a19090783 - 10 Sep 2026

Abstract

The increasing penetration of electric vehicles (EVs) creates new challenges for coordinated planning of charging and battery-swapping infrastructure. This study aims to develop a joint location and capacity planning framework for electric vehicle charging stations (EVCSs) and electric vehicle swapping stations (EVSSs) in [...] Read more.

The increasing penetration of electric vehicles (EVs) creates new challenges for coordinated planning of charging and battery-swapping infrastructure. This study aims to develop a joint location and capacity planning framework for electric vehicle charging stations (EVCSs) and electric vehicle swapping stations (EVSSs) in an electric–traffic coupled system, considering infrastructure cost, user service requirements, and distribution-network performance. A multi-objective planning model is established based on spatial-temporal energy replenishment demand and traffic–grid coupling. To improve computational efficiency, a random forest surrogate-assisted NSGA-III (RF-SA-NSGA-III) is proposed, in which expensive user-service and voltage-related objective evaluations are selectively approximated by random forest models, combined with periodic true-model correction and final verification. Case studies show that the proposed method obtains competitive Pareto-optimal solutions and replaces 89.38% of expensive evaluations, reducing computational time from 45,706.3 s to 4689.4 s. The framework provides practical support for coordinated EVCS–EVSS siting and capacity allocation while balancing investment, user service quality, and voltage performance. Full article

(This article belongs to the Special Issue Transportation and Traffic Engineering)

► Show Figures

Figure 1

36 pages, 1825 KB

Open AccessArticle

Multi-Objective Optimization of Resource-Constrained Construction Scheduling for Power Transmission and Distribution Projects Considering Prefabricated Components

by Yang Bai, Xiangyong An, Haibo Zhao, Xiao Fan, Xiying Fan, Tingjun Li, Zhuowen Zuo and Xiaoqing Han

Algorithms 2026, 19(9), 782; https://doi.org/10.3390/a19090782 - 10 Sep 2026

Abstract

Research on low-carbon construction in power transmission and distribution (PTD) projects has mainly focused on carbon accounting and emission-reduction assessment, with limited consideration of construction method differences and shared resource constraints. This study proposes a multi-objective optimization method for resource-constrained construction scheduling considering [...] Read more.

Research on low-carbon construction in power transmission and distribution (PTD) projects has mainly focused on carbon accounting and emission-reduction assessment, with limited consideration of construction method differences and shared resource constraints. This study proposes a multi-objective optimization method for resource-constrained construction scheduling considering prefabricated components. The project is decomposed into construction units and procedures, and candidate construction methods are defined by carbon emissions, cost, duration, prefabrication rate, and resource demand. The lower level generates representative schemes from procedure-level method combinations, while the upper-level mixed-integer linear programming (MILP) model determines construction-unit schemes, resource input levels, and start times under precedence and shared resource constraints. An ε-constraint method is used to obtain a carbon-prioritized scheme. A 110 kV PTD case study shows that, compared with the cast-in-place baseline, the carbon-prioritized scheme reduces carbon emissions and duration by 17.45% and 14.66%, respectively, while maintaining comparable cost. Compared with the minimum-carbon scheme, it increases carbon emissions by only 0.21%, but shortens duration by 10.00% and slightly reduces cost. The results demonstrate that the proposed method can shorten construction duration and reduce cost while keeping carbon emissions close to the minimum level. Full article

23 pages, 4072 KB

Open AccessArticle

AccreditNet: An Explainable Data-Adaptive AI Framework for Accreditation Quality Analytics and Decision Support in the Training Sector

by Fahd M. Aldosari, Donia Y. Badawood and Khaled H. Almotairi

Algorithms 2026, 19(9), 781; https://doi.org/10.3390/a19090781 - 10 Sep 2026

Abstract

Accreditation and quality assurance in technical and vocational education and training (TVET) remain largely dependent on manual review and expert judgment. This study presents AccreditNet, an explainable tabular AI model embedded in a broader accreditation decision support framework. The experimentally evaluated component combines [...] Read more.

Accreditation and quality assurance in technical and vocational education and training (TVET) remain largely dependent on manual review and expert judgment. This study presents AccreditNet, an explainable tabular AI model embedded in a broader accreditation decision support framework. The experimentally evaluated component combines per-feature tokenization, multi-head self-attention, hierarchical criterion fusion, class-balanced focal loss, integrated gradients, and descriptive performance gap scoring. Evaluation is conducted against a structured NAAC benchmark that includes criterion scores and institutional attributes. Because the official NAAC grade is derived from criterion-level scoring, the benchmark is interpreted as a test of how well the model represents the criterion-to-grade mapping rather than as evidence of an independent predictive relationship. On the reported held-out split, AccreditNet achieved 93.72% accuracy, 92.84% macro F1-score, and 0.9706 AUC-ROC. OCR-based ingestion, NLP/LLM-assisted evidence mapping, automated reporting, and Saudi deployment are proposed operational extensions and were not evaluated in the present experiments. The results support further study of explainable accreditation analytics and decision support, subject to validation on local data. Full article

(This article belongs to the Section Algorithms for Multidisciplinary Applications)

► Show Figures

Figure 1

29 pages, 9912 KB

Open AccessArticle

An Importance Sampling Monte Carlo Framework for Fan Vote Reconstruction and Voting Rule Evaluation: A Case Study of Dancing with the Stars

by Renrui Han, Yufan Sun, Siyuan Yin, Jingxuan Wang, Jintian Ji and Juntong Liu

Algorithms 2026, 19(9), 780; https://doi.org/10.3390/a19090780 - 9 Sep 2026

Abstract

In television dance competitions such as Dancing with the Stars (DWTS), fair and transparent outcome determination is crucial for addressing public concerns regarding contest fairness. However, existing studies face challenges due to the opacity of fan voting and the lack of a systematic [...] Read more.

In television dance competitions such as Dancing with the Stars (DWTS), fair and transparent outcome determination is crucial for addressing public concerns regarding contest fairness. However, existing studies face challenges due to the opacity of fan voting and the lack of a systematic evaluation of rule changes, making fairness disputes difficult to resolve. To address these challenges, this study develops a data-driven counterfactual analysis framework to quantitatively assess the impact of different voting rules on competition outcomes. We develop a constraint-based Monte Carlo framework to characterize feasible fan-vote distributions consistent with historical elimination results, together with an IS-MC variant for controlled reconstruction comparisons. A paired semi-synthetic validation comprising 1145 trials across five fan-vote-generating mechanisms, with the generated vote shares concealed during reconstruction, showed that IS-MC achieved the lowest mean MSE (0.01915) and MAE (0.08980), together with the highest mean Spearman correlation (0.382) and empirical 95% interval coverage (0.879) among the compared methods under a common sampling budget. Its lower ESS-to-valid-sample ratio nevertheless suggests a trade-off between reconstruction quality and importance-weight stability. Within the modeled counterfactual comparisons, the percentage-based rule was more sensitive to variation in fan-vote proportions, and contestants whose reconstructed fan support was high relative to their judge-score shares tended to receive more favorable modeled rankings under that rule. Under the stylized bottom-two judge-selection scenario, the modeled outcomes differed when judges retained the contestant with the higher judge score; this is a conditional sensitivity finding rather than evidence about actual judges’ decisions. Additionally, a six-feature GBDT proxy was evaluated using chronologically held-out percentage-rule observations, and SHAP analysis on the predicted-probability scale was used to characterize feature associations with modeled counterfactual reversals. We construct an end-to-end system encompassing data preprocessing, model inversion, uncertainty quantification, semi-synthetic validation, and diagnostic visualization, providing a descriptive basis for comparing DWTS voting rules. The proposed methodological framework can also be extended to other competitive scenarios relying on hybrid decision-making. Full article

(This article belongs to the Topic Green Computing and Artificial Intelligence: Energy- and Resource-Efficient Intelligent Systems)

► Show Figures

Figure 1

29 pages, 12641 KB

Open AccessArticle

Traffic-Flow Prediction Using T-AGCN Based on an Adaptive Search Graph Fusion Mechanism

by Hai Yan, Rui Jian, Chengcheng Wang and Jianwen Yao

Algorithms 2026, 19(9), 779; https://doi.org/10.3390/a19090779 - 9 Sep 2026

Abstract

The rapid growth in traffic volumes has increased the demand for traffic-flow forecasting models with stronger prediction capability. Traditional methods that rely on local feature extraction and static spatial graph construction can no longer fully meet these requirements. To address the short- and [...] Read more.

The rapid growth in traffic volumes has increased the demand for traffic-flow forecasting models with stronger prediction capability. Traditional methods that rely on local feature extraction and static spatial graph construction can no longer fully meet these requirements. To address the short- and long-term fluctuations in freeway traffic flow and the dynamic correlations among node flows, this study proposes ASGF-TAGCN, a traffic-flow forecasting model based on a component-specific multi-source graph fusion mechanism. The model integrates CEEMDAN-based multiscale decomposition, multi-source graph construction (physical topology, DTW-based semantic similarity, and node-adaptive learning), a learnable fusion mechanism for multi-source spatial modeling, Transformer-based temporal modeling, and IHPO hyperparameter optimization. Across three independent training runs at the 30 min forecasting horizon, ASGF-TAGCN achieved a mean MAE of 2.2982 ± 0.1181 and a mean RMSE of 3.1147 ± 0.1059, where the variability is reported as the sample standard deviation. In the fixed seed-42 run used for the baseline comparison, ASGF-TAGCN achieved an MAE of 2.219 and an RMSE of 3.063, reducing the two errors by 14.72% and 11.58%, respectively, relative to T-AGCN, the strongest baseline among the selected models. These results demonstrate that ASGF-TAGCN consistently reduces prediction errors and effectively captures complex spatiotemporal dependencies, offering a reliable solution for short-term freeway traffic-flow forecasting. Full article

(This article belongs to the Special Issue Algorithms for Smart Cities (3rd Edition))

► Show Figures

Figure 1

35 pages, 4896 KB

Open AccessArticle

Domain-Adaptive Audio Large Language Model for Acoustic Fault Diagnosis and Semantic Description of Coal Mine Equipment

by Daming Cui, Xin Zhang and Qiang Ma

Algorithms 2026, 19(9), 778; https://doi.org/10.3390/a19090778 - 9 Sep 2026

Abstract

Underground coal-mine equipment operates under broadband noise, high dust, humidity, and methane. Acoustic sensing is uniquely suited to this environment: it captures vibration, friction, and airflow signatures without physical contact, incurs low sensor-deployment cost, responds at millisecond speed, and remains effective in low-light, [...] Read more.

Underground coal-mine equipment operates under broadband noise, high dust, humidity, and methane. Acoustic sensing is uniquely suited to this environment: it captures vibration, friction, and airflow signatures without physical contact, incurs low sensor-deployment cost, responds at millisecond speed, and remains effective in low-light, high-dust conditions where optical and vibration alternatives fail. Acoustic fault perception is therefore critically important in underground coal-mine operations. Three unresolved challenges remain: (i) motor whine, material-collision impacts, and ventilation-fan roar compound into a low-SNR soundscape where conventional models lose noise robustness; (ii) acoustic signatures vary widely across equipment types and fault-development stages; and (iii) existing supervised classifiers, trained on imbalanced data, exhibit limited generalization and output only binary judgments, lacking the semantic descriptions that maintenance crews actually need. To address all three, we propose a domain-adaptive audio LLM coupling a BEATs encoder (frozen during Stage II, adapted via Low-Rank Adaptation (LoRA) during Stage I), a Querying Transformer (Q-Former) alignment layer with Dynamic Acoustic Token Compression (DATC), a LLaMA-3.1-8B decoder adapted via LoRA, and Constrained Decoding for Structured Fault Description (CD-SFD) enforcing a three-slot output of fault type, danger level, and handling recommendation. DATC allocates query budget by signal energy to suppress noise-dominated frames; CD-SFD is a state-machine decoder that guarantees the three-slot schema. We release CMEASD: 1200 recordings comprising 24 physical machines (4 per equipment type, 12/6/6 machine-disjoint split). Under machine-disjoint evaluation, the model reaches Macro Accuracy 84.6 ± 1.5% and Macro F1 82.9 ± 1.6%, outperforming the strongest discriminative baseline (PANNs-Transformer, 81.7%) by +2.9 pp and SALMONN-LoRA by +2.5 pp. Ablations attribute +2.4/+1.4/+0.9 pp to DATC, CD-SFD, and the domain prompt. A 30-day mine trial achieves 7.0 s end-to-end latency with 21/30 days of stable, zero-false-shutdown operation. Full article

(This article belongs to the Special Issue Deep Learning Methods and Applications)

► Show Figures

Figure 1

21 pages, 2395 KB

Open AccessArticle

Design of Automatic Short-Answer Scoring Prediction Model in Academic Courses Based on Statistical and Deep Learning Algorithms

by Muhammad Umar Farooq, Tauqir Ahmad and Muhammad Aslam

Algorithms 2026, 19(9), 777; https://doi.org/10.3390/a19090777 - 9 Sep 2026

Abstract

In the digital education system, the increasing demand for measuring lexical and semantic text similarity has led to the need for developing scalable and intelligent automated grading systems. Traditional assessment methods used for short-answer grading are labor-intensive, time-consuming and prone to human-evaluator bias. [...] Read more.

In the digital education system, the increasing demand for measuring lexical and semantic text similarity has led to the need for developing scalable and intelligent automated grading systems. Traditional assessment methods used for short-answer grading are labor-intensive, time-consuming and prone to human-evaluator bias. Furthermore, computer science curricula have different evaluation challenges as programming-centric courses require rigorous syntactical and structural validation, whereas theoretical courses require deep conceptual and semantic comprehension. To address these dual challenges, we introduce the Virtual University Automatic Short-Answer Grading (VUASAG) framework, a model addressing both statistical-based lexical analysis and Transformer-based semantic architectures. Utilizing an expanded Mohler dataset spanning multiple computer science domains (including Data Structures, Introduction to Programming, Object-Oriented Programming, and Software Engineering), we evaluated the efficiency of traditional string-matching metrics alongside state-of-the-art Transformer-based models (T5, BERT, XLNet, and SBERT). The empirical results demonstrate that token-based statistical models perform robustly on keyword-dependent syntax validation while Transformer-based models excel by capturing deep conceptual context. Specifically, Sentence-BERT (SBERT) achieved superior predictive accuracy across courses, yielding a minimum Root Mean Square Error (RMSE) of 0.9511 for the Introduction to Programming course, paired with moderate positive Pearson and Spearman correlation coefficients. Confidence intervals are estimated using Fisher’s z-transformation with a 95 percentile confidence level. Question-level five-fold and two-fold cross-validation is conducted for all four courses, and SBERT is found to perform better. Finally, to demonstrate the practical utility, we deployed a web-based portal for students’ evaluation leveraging a LLaMA-based pre-trained model to deliver real-time, personalized pedagogical feedback measuring code functionality, quality, documentation, and error handling. This reduces the educator workload while preserving grading integrity. Full article

(This article belongs to the Topic Advancements in Coupling Simulation and Artificial Intelligence: Methods, Applications, and Future Directions)

► Show Figures

Figure 1

12 pages, 1437 KB

Open AccessArticle

A Reinforcement Learning-Based Scheduling Algorithm for Special Material Transportation

by Jianbo Zhao and Xiang Su

Algorithms 2026, 19(9), 776; https://doi.org/10.3390/a19090776 - 9 Sep 2026

Abstract

Efficient scheduling of special materials is essential for improving the operational efficiency of material transportation systems. However, multiple processing stages, heterogeneous transportation resources, and complex transfer paths pose significant challenges to efficient scheduling. To address these challenges, this paper proposes an improved reinforcement [...] Read more.

Efficient scheduling of special materials is essential for improving the operational efficiency of material transportation systems. However, multiple processing stages, heterogeneous transportation resources, and complex transfer paths pose significant challenges to efficient scheduling. To address these challenges, this paper proposes an improved reinforcement learning (RL)-based scheduling algorithm. First, a scheduling optimization model is established with the objectives of minimizing makespan and balancing resource utilization. Second, an improved action-value update strategy is developed to reduce Q-value estimation bias, thereby improving policy convergence and scheduling performance. Experimental results show that the proposed Improved DQN outperforms greedy, genetic, and standard DQN algorithms across different task and resource scales. In particular, it achieves an average relative error rate of 27.48% with 500 tasks and a maximum error rate of only 6.51% under different resource configurations, demonstrating its effectiveness and scalability for complex special material scheduling. Full article

(This article belongs to the Section Evolutionary Algorithms and Machine Learning)

► Show Figures

Figure 1

15 pages, 1882 KB

Open AccessArticle

Automatic Modulation Recognition Based on RadioAlign Feature Adaptation, Scale Attention, and a Bayesian Residual Branch

by Aiming Zhang, Yongchun Liu and Yawen Lei

Algorithms 2026, 19(9), 775; https://doi.org/10.3390/a19090775 - 9 Sep 2026

Abstract

To address amplitude scaling, phase rotation, local misalignment, and ambiguous decision boundaries under low-SNR and non-cooperative reception conditions, we propose an Attention RadioAlign Bayesian Residual Network (ABRNet) for automatic modulation recognition. ABRNet takes raw I/Q sequences as input. It first employs a RadioAlign [...] Read more.

To address amplitude scaling, phase rotation, local misalignment, and ambiguous decision boundaries under low-SNR and non-cooperative reception conditions, we propose an Attention RadioAlign Bayesian Residual Network (ABRNet) for automatic modulation recognition. ABRNet takes raw I/Q sequences as input. It first employs a RadioAlign module to estimate and apply bounded learnable transformations that aim to mitigate amplitude-phase and local-offset input variations. A scale-attention module then adaptively weights short-term phase transitions, local waveform patterns, and long-term envelope variations. Finally, a Bayesian residual branch adds a small stochastic residual to the deterministic logits and estimates the predictive mean, variance, and entropy as reliability measures. Across five random seeds, ABRNet achieves a mean overall accuracy of 63.4668% (sample standard deviation: 0.0583%), a low-SNR accuracy of 34.2864% (0.1399%), a high-SNR accuracy of 92.6473% (0.0630%), and a Macro F1 score of 65.7086% (0.0784%). The error-detection AUROC is 0.894092 (0.000370), the expected calibration error is 0.016618 (0.001703), and the Brier score is 0.434423 (0.000127). With approximately 287 K parameters, ABRNet provides classification and predictive reliability measures in an end-to-end I/Q processing pipeline. Full article

► Show Figures

Figure 1

27 pages, 6835 KB

Open AccessArticle

A Reliability-Aware Cross-Branch Contrastive Graph Convolutional Network for Hyperspectral Image Processing

by Runhao Zhang, Wanzhang Wang, Wei Feng, Fei Yu and Haize Hu

Algorithms 2026, 19(9), 774; https://doi.org/10.3390/a19090774 - 9 Sep 2026

Abstract

Hyperspectral images contain abundant spectral information and provide fine-grained spatial representations. However, their high dimensionality, severe spectral redundancy, subtle inter-class differences, mixed boundary regions, and limited labeled samples pose significant challenges to accurate classification. Convolutional neural networks (CNNs) have limited capability in modeling [...] Read more.

Hyperspectral images contain abundant spectral information and provide fine-grained spatial representations. However, their high dimensionality, severe spectral redundancy, subtle inter-class differences, mixed boundary regions, and limited labeled samples pose significant challenges to accurate classification. Convolutional neural networks (CNNs) have limited capability in modeling non-Euclidean structural relationships, whereas graph convolutional networks (GCNs) are susceptible to the quality of superpixel segmentation and noise propagation over graph structures. To address these issues in hyperspectral image classification, this paper proposes a Reliability-Aware Cross-Branch Contrastive Graph Convolutional Network (RACB-CGCN). The proposed method employs a dual-branch CNN–GCN architecture to extract pixel-level local spectral–spatial features and superpixel-level structural features, respectively. A superpixel reliability estimation and propagation control mechanism is introduced to assess node reliability based on the discrepancy between pixel-level features and superpixel-reconstructed features. This mechanism effectively suppresses the propagation of noisy information caused by impure superpixels and mixed boundary regions. Meanwhile, a cross-branch supervised contrastive learning strategy is developed to enhance semantic consistency between the CNN and GCN branches, thereby improving intra-class compactness and inter-class separability. In addition, a class-adaptive fusion module is designed to dynamically adjust the contributions of the two branches according to the feature characteristics of different land-cover classes. Experimental results demonstrate that the proposed method effectively exploits the complementary information between pixel-level fine-grained features and superpixel-level structural features, leading to improved classification accuracy and robustness in hyperspectral image classification. Full article

► Show Figures

Figure 1

Journal Description

Algorithms

Latest Articles

Journal Menu

Journal Browser

Highly Accessed Articles

Latest Books

E-Mail Alert

News

Topics

Conferences

Special Issues

Topical Collections

Further Information

Guidelines

MDPI Initiatives

Follow MDPI