Multi-Stage Dual-Perturbation Attack Targeting Transductive SVMs and the Corresponding Adversarial Training Defense Mechanism DOI Open Access
Li Liu, Haiyan Chen, Changchun Yin

et al.

Electronics, Journal Year: 2024, Volume and Issue: 13(24), P. 4984 - 4984

Published: Dec. 18, 2024

The Transductive Support Vector Machine (TSVM) is an effective semi-supervised learning algorithm vulnerable to adversarial sample attacks. This paper proposes a new attack method called the Multi-Stage Dual-Perturbation Attack (MSDPA), specifically targeted at TSVMs. MSDPA has two phases: initial samples are generated by arbitrary range attack, and finer attacks performed on critical features induce TSVM generate false predictions. To improve TSVM’s defense against MSDPAs, we incorporate training into loss function minimize of both standard during process. improved considers samples’ effect enhances model’s robustness. Experimental results several datasets show that our proposed defense-enhanced (adv-TSVM) performs better in classification accuracy robustness than native other baseline algorithms, such as S3VM. study provides solution capability kernel methods setting.

Language: Английский

An explainable artificial-intelligence-aided safety factor prediction of road embankments DOI Creative Commons
Azam Abdollahi, Deli Li, Jian Deng

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2024, Volume and Issue: 136, P. 108854 - 108854

Published: July 4, 2024

Despite the widespread application of data-centric techniques in Geotechnical Engineering, there is a rising need for building trust artificial intelligence (AI)-driven safety assessment road embankments due to its so-called "black-box" nature. In addition, from lens limit equilibrium approaches, e.g., Bishop, Fellenius, Janbu and Morgenstern–Price, finite element method, it essential carefully examine interplay both topological physical/mechanical properties during factor (FoS) predictions. First, aside having conventional geotechnical inputs soil core foundation height embankments, this paper codifies geometric features innovatively. The number slope types with different ratios including 1:1, 1.5:1 2:1 as well berms introduced. Second, pool 19 machine learning (ML) effortlessly trained on dataset using an automated ML (AutoML) pipeline identify most optimized algorithm. Finally, achieve post-hoc interpretability internal mechanism input–output relationship unbiasedly, game-theory-based explainable AI (XAI) method called Shapley additive explanations (SHAP) values applied. SHAP-aided importance analysis provides human-interpretable insights indicates height, California bearing ratio, type cohesion influential parameters. Exclusively, analyzing hazardous by classifying main joint contributors exhibits complex highly variable influence FoS. This harnesses power XAI tools enhance reliability transparency rapid FoS prediction slopes. It targets researchers, practitioners, decision-makers, general public first time.

Language: Английский

Citations

16
Imbalanced rock burst assessment using variational autoencoder-enhanced gradient boosting algorithms and explainability DOI Creative Commons
Shan Lin,

Zenglong Liang,

Miao Dong

et al.

Underground Space, Journal Year: 2024, Volume and Issue: 17, P. 226 - 245

Published: Jan. 21, 2024

We conducted a study to evaluate the potential and robustness of gradient boosting algorithms in rock burst assessment, established variational autoencoder (VAE) address imbalance dataset, proposed multilevel explainable artificial intelligence (XAI) tailored for tree-based ensemble learning. collected 537 data from real-world records selected four critical features contributing occurrences. Initially, we employed visualization gain insight into data's structure performed correlation analysis explore distribution feature relationships. Then, set up VAE model generate samples minority class due imbalanced distribution. In conjunction with VAE, compared evaluated six state-of-the-art models, including classical logistic regression model, prediction. The results indicated that outperformed single VAE-classifier original classifier, VAE-NGBoost yielding most favorable results. Compared other resampling methods combined NGBoost datasets, such as synthetic oversampling technique (SMOTE), SMOTE-edited nearest neighbours (SMOTE-ENN), SMOTE-tomek links (SMOTE-Tomek), yielded best performance. Finally, developed XAI using sensitivity analysis, Tree Shapley Additive exPlanations (Tree SHAP), Anchor provide an in-depth exploration decision-making mechanics VAE-NGBoost, further enhancing accountability models predicting

Language: Английский

Citations

8
A Machine Learning-Based Electricity Consumption Forecast and Management System for Renewable Energy Communities DOI Creative Commons

Miguel Matos,

João Almeida, Pedro Gonçalves

et al.

Energies, Journal Year: 2024, Volume and Issue: 17(3), P. 630 - 630

Published: Jan. 28, 2024

The energy sector is currently undergoing a significant shift, driven by the growing integration of renewable sources and decentralization electricity markets, which are now extending into local communities. This transformation highlights pivotal role prosumers within these as result, concept Renewable Energy Communities gaining traction, empowering their members to curtail reliance on non-renewable facilitating generation, storage, exchange. Also in community, management efficiency depends being able predict future consumption make decisions regarding purchase, sale storage electricity, why forecasting community extremely important. study presents an innovative approach manage balance, relying Machine Learning (ML) techniques, namely eXtreme Gradient Boosting (XGBoost), forecast consumption. Subsequently, decision algorithm employed for trading with public grid, based solar production forecasts, levels market prices. outcomes simulated model demonstrate efficacy incorporating since system showcases potential reduce both expenses its dependence from centralized distribution grid. ML-based techniques allowed better results specially bi-hourly tariffs high capacity scenarios bill reductions 9.8%, 2.8% 5.4% high, low, average photovoltaic (PV) generation levels, respectively.

Language: Английский

Citations

6
Evaluating machine learning model for investigating surface chloride concentration of concrete exposed to tidal environment DOI
Thi Tuyet Trinh Nguyen,

Long Khanh Nguyen

Frontiers of Structural and Civil Engineering, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 22, 2025

Language: Английский

Citations

0
Interpretable artificial intelligence for advancing slope stability assessment techniques with Technosols DOI
Jiazhou Li, Mengjie Huang, M Ma

et al.

Soil Use and Management, Journal Year: 2025, Volume and Issue: 41(1)

Published: Jan. 1, 2025

Abstract Slope stability is a critical factor in ensuring the safety and longevity of infrastructure, especially areas prone to landslides soil erosion. Traditional methods slope assessment, while widely used, often struggle provide accurate results when applied Technosols—soils modified by human activities composed waste materials. This study proposes novel approach that combines artificial intelligence techniques improve precision predictions these complex types. The method utilizes model based on neural networks, trained large dataset factors. Unlike conventional techniques, proposed integrates multiple environmental material properties more assessment compared other models. model's performance demonstrated R 2 values .999975 for test datasets, which significantly better than similar work statistical analysis. Moreover, incorporating Shapley Additive Explanations (SHAP), we clear understanding impact various parameters stability. findings suggest machine learning‐based offers reliable tool evaluation Technosols, making it valuable addition field.

Language: Английский

Citations

0
Rock Slope Stability Prediction: A Review of Machine Learning Techniques DOI

Arifuggaman Arif,

Chunlei Zhang,

Mahabub Hasan Sajib

et al.

Geotechnical and Geological Engineering, Journal Year: 2025, Volume and Issue: 43(3)

Published: Feb. 18, 2025

Language: Английский

Citations

0
Random forest-based prediction of shallow slope stability considering spatiotemporal variations in unsaturated soil moisture DOI Creative Commons
Yangyang Li,

Saranya Rangarajan,

Y.M. Cheng

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 13, 2025

With the increasing incidence of extreme rainfall driven by global climate change, geological hazards like landslides have become more prevalent. This study proposed an efficient framework that combined machine learning and physical models to enhance computational efficiency reliability for regional slope stability predictions under rainfall. The GEOtop model was employed simulate volumetric water content (VWC) in unsaturated soil area Singapore maximum daily 5-day antecedent conditions. result analyses then incorporated into Scoops3D factor safety (FOS) calculations. random forest (RF) were trained using VWC applied predict rainfall, with outcomes compared those Scoops3D. Statistical results spatial distribution maps both showed achieved comparable accuracy at various depths while significantly improving efficiency. findings also highlighted critical role surface moisture (at 0.05 m) predictions. demonstrates potential integrating prediction, as well supports integration remote sensing or field-measured data dynamic

Language: Английский

Citations

0
Application of machine learning in early warning system of geotechnical disaster: a systematic and comprehensive review DOI Creative Commons
Shan Lin,

Zenglong Liang,

Hongwei Guo

et al.

Artificial Intelligence Review, Journal Year: 2025, Volume and Issue: 58(6)

Published: March 17, 2025

Language: Английский

Citations

0
Predicting the sustainability of e-waste mortar for mitigating thermal spalling cracks using ANN and RSM DOI
Yacine Abadou, Abderrahmane Ghrieb,

T. Choungara

et al.

Innovative Infrastructure Solutions, Journal Year: 2025, Volume and Issue: 10(4)

Published: March 27, 2025

Language: Английский

Citations

0
A FEM-guided data-driven machine learning model for residual stress characterization in ultrasonic surface rolling of lightweight alloys DOI

Rahul Pradhan,

Farag M. A. Altalbawy, Ahmed Raza Khan

et al.

Applied Physics A, Journal Year: 2024, Volume and Issue: 130(6)

Published: May 14, 2024

Language: Английский

Citations

3