Future of machine learning in geotechnics

Georisk Assessment and Management of Risk for Engineered Systems and Geohazards, Journal Year: 2022, Volume and Issue: 17(1), P. 7 - 22

Published: June 16, 2022

Machine learning (ML) is widely used in many industries, resulting recent interests to explore ML geotechnical engineering. Past review papers focus mainly on algorithms while this paper advocates an agenda put data at the core, develop novel that are effective for (existing and new), address needs of current practice, exploit new opportunities from emerging technologies or meet digital transformation, take advantage knowledge accumulated experience. This called data-centric geotechnics it contains three core elements: centricity, fit (and transform) context. The future machine should be envisioned with “data first practice central” mind. Data-driven site characterization (DDSC) active research topic because understanding ground crucial all projects. Examples DDSC challenges ugly explainable recognition. Additional include making indispensable (ML supremacy), how learn (meta-learning), becoming smart (digital twin).

Language: Английский

---

A survey of machine learning and deep learning in remote sensing of geological environment: Challenges, advances, and opportunities

ISPRS Journal of Photogrammetry and Remote Sensing, Journal Year: 2023, Volume and Issue: 202, P. 87 - 113

Published: June 13, 2023

Language: Английский

---

A comprehensive comparison among metaheuristics (MHs) for geohazard modeling using machine learning: Insights from a case study of landslide displacement prediction

Engineering Applications of Artificial Intelligence, Journal Year: 2022, Volume and Issue: 114, P. 105150 - 105150

Published: July 7, 2022

Machine learning (ML) has been extensively applied to model geohazards, yielding tremendous success. However, researchers and practitioners still face challenges in enhancing the reliability of ML models. In present study, a systematic framework combining k-fold cross-validation (CV), metaheuristics (MHs), support vector regression (SVR), Friedman Nemenyi tests was proposed improve performance geohazard modeling. The average normalized mean square error (NMSE) from CV sets adopted as fitness metric. Twenty most well-established MHs recent were tune hyperparameters SVR evaluated through nonparametric post hoc identify significant differences. Observations typical reservoir landslide selected benchmark dataset, accuracy, robustness, computational time, convergence speed compared. Significant differences among twenty identified by absolute (MAE), root squared (RMSE), Kling–Gupta efficiency (KGE), with p values lower than 0.05. comparison results demonstrated that multiverse optimizer (MVO) is highest-performing, stable, computationally efficient algorithms, providing superior other methods, nearly optimum correlation coefficient (R), low MAE (23.5086 versus 23.9360), RMSE (48.6946 50.1882), high KGE (0.9803 0.9893) predicting displacement Shuping landslide. This paper considerably enriches literature regarding hyperparameter optimization algorithms enhancement their reliability. addition, have potential for evaluating comparing various ML-based

Language: Английский

---

Metaheuristic-based support vector regression for landslide displacement prediction: a comparative study

Landslides, Journal Year: 2022, Volume and Issue: 19(10), P. 2489 - 2511

Published: June 30, 2022

Abstract Recently, integrated machine learning (ML) metaheuristic algorithms, such as the artificial bee colony (ABC) algorithm, genetic algorithm (GA), gray wolf optimization (GWO) particle swarm (PSO) and water cycle (WCA), have become predominant approaches for landslide displacement prediction. However, these algorithms suffer from poor reproducibility across replicate cases. In this study, a hybrid approach integrating k-fold cross validation (CV), support vector regression (SVR), nonparametric Friedman test is proposed to enhance reproducibility. The five previously mentioned metaheuristics were compared in terms of accuracy, computational time, robustness, convergence. results obtained Shuping Baishuihe landslides demonstrate that can be utilized determine optimum hyperparameters present statistical significance, thus enhancing accuracy reliability ML-based Significant differences observed among metaheuristics. Based on test, which was performed root mean square error (RMSE), Kling-Gupta efficiency (KGE), PSO recommended hyperparameter tuning SVR-based prediction due its ability maintain balance between precision, robustness. promising presenting

Language: Английский

---

Deep learning methods for time-dependent reliability analysis of reservoir slopes in spatially variable soils

Computers and Geotechnics, Journal Year: 2023, Volume and Issue: 159, P. 105413 - 105413

Published: March 28, 2023

Language: Английский

---

Modelling landslide susceptibility prediction: A review and construction of semi-supervised imbalanced theory

Earth-Science Reviews, Journal Year: 2024, Volume and Issue: 250, P. 104700 - 104700

Published: Jan. 29, 2024

Language: Английский

---

Iterative integration of deep learning in hybrid Earth surface system modelling

Nature Reviews Earth & Environment, Journal Year: 2023, Volume and Issue: 4(8), P. 568 - 581

Published: July 11, 2023

Language: Английский

---

Predictive deep learning for pitting corrosion modeling in buried transmission pipelines

Process Safety and Environmental Protection, Journal Year: 2023, Volume and Issue: 174, P. 320 - 327

Published: April 5, 2023

Despite significant efforts and investments in the renewable energy sector, fossil fuels continue to provide majority of world's supply. Transmission pipelines, which are extensively used oil gas industry, vulnerable various failure mechanisms, such as corrosion. Among these, pitting corrosion offshore pipelines is most prevalent type external This study explores potential deep learning models (Generalization Generalization-Memorization models) predict maximum depth pipelines. The trained considering characteristics soil where pipe buried different types protective coating pipes. application neural networks resulted a mean squared error prediction 0.0055 training data 0.0037 test data. These results demonstrate that outperform all empirical hybrid applied previous studies on same dataset. proposed model this has rates due enhance safety reliability these facilities.

Language: Английский

---

A systematic review of trustworthy artificial intelligence applications in natural disasters

Computers & Electrical Engineering, Journal Year: 2024, Volume and Issue: 118, P. 109409 - 109409

Published: June 29, 2024

Artificial intelligence (AI) holds significant promise for advancing natural disaster management through the use of predictive models that analyze extensive datasets, identify patterns, and forecast potential disasters. These facilitate proactive measures such as early warning systems (EWSs), evacuation planning, resource allocation, addressing substantial challenges associated with This study offers a comprehensive exploration trustworthy AI applications in disasters, encompassing management, risk assessment, prediction. research is underpinned by an review reputable sources, including Science Direct (SD), Scopus, IEEE Xplore (IEEE), Web (WoS). Three queries were formulated to retrieve 981 papers from earliest documented scientific production until February 2024. After meticulous screening, deduplication, application inclusion exclusion criteria, 108 studies included quantitative synthesis. provides specific taxonomy disasters explores motivations, challenges, recommendations, limitations recent advancements. It also overview techniques developments using explainable artificial (XAI), data fusion, mining, machine learning (ML), deep (DL), fuzzy logic, multicriteria decision-making (MCDM). systematic contribution addresses seven open issues critical solutions essential insights, laying groundwork various future works trustworthiness AI-based management. Despite benefits, persist In these contexts, this identifies several unused used areas disaster-based theory, collects ML, DL techniques, valuable XAI approach unravel complex relationships dynamics involved utilization fusion processes related Finally, extensively analyzed ethical considerations, bias, consequences AI.

Language: Английский

---

A review of remote sensing image segmentation by deep learning methods

International Journal of Digital Earth, Journal Year: 2024, Volume and Issue: 17(1)

Published: March 18, 2024

Remote sensing (RS) images enable high-resolution information collection from complex ground objects and are increasingly utilized in the earth observation research. Recently, RS technologies continuously enhanced by various characterized platforms sensors. Simultaneously, artificial intelligence vision algorithms also developing vigorously playing a significant role image analysis. In particular, aiming to divide into different elements with specific semantic labels, segmentation could realize visual acquisition interpretation. As one of pioneering methods advantages deep feature extraction ability, learning (DL) have been exploited proved be highly beneficial for precise recent years. this paper, comprehensive review is performed on remote survey systems kinds specially designed architectures. Meanwhile, DL-based applied four domains illustrated, including geography, precision agriculture, hydrology, environmental protection issues. end, existing challenges promising research directions discussed. It envisioned that able provide technical reference, deployment successful exploitation DL empowered approaches.

Language: Английский

---