Prediction of Thermal Diffusivity of Non-Plastic Soil for the Design of Ground Heat Exchanger Using Machine Learning Approach

Indian geotechnical journal, Journal Year: 2025, Volume and Issue: unknown

Published: May 1, 2025

Language: Английский

---

The Diverse Mountainous Landslide Dataset (DMLD): A High-Resolution Remote Sensing Landslide Dataset in Diverse Mountainous Regions

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(11), P. 1886 - 1886

Published: May 24, 2024

The frequent occurrence of landslides poses a serious threat to people’s lives and property. In order evaluate disaster hazards based on remote sensing images via machine learning means, it is essential establish an image database with manually labeled boundaries landslides. However, the existing datasets do not cover diverse types mountainous To address this issue, we propose high-resolution (1 m) landslide dataset (DMLD), including 990 instances across different terrain in southwestern China. performance DMLD, seven state-of-the-art deep models loss functions were implemented it. experiment results demonstrate only that all these methods characteristics can adapt well but also DMLD has potential adaptability other geographical regions.

Language: Английский

---

Experimental study on dynamic response and failure mode of bedding rock slope with cracks under earthquake

Bulletin of Engineering Geology and the Environment, Journal Year: 2025, Volume and Issue: 84(2)

Published: Jan. 24, 2025

Language: Английский

---

Mesoscale investigation on transport law in nylon porous medium based on the lattice Boltzmann methodology

Colloids and Surfaces A Physicochemical and Engineering Aspects, Journal Year: 2025, Volume and Issue: 711, P. 136370 - 136370

Published: Feb. 6, 2025

Language: Английский

---

Machine Learning-Based Model for Prediction Permeability in Porous Media: Method and Application to Unconventional Reservoirs

International Petroleum Technology Conference, Journal Year: 2025, Volume and Issue: unknown

Published: Feb. 17, 2025

Abstract Accurately predicting permeability in porous media is crucial for various engineering fields, including petroleum engineering, geology, and environmental science. Unlike conventional reservoirs, shale reservoirs predominantly feature micro- to nano-scale pores, making prediction challenging difficult obtain through experimental methods. This research presents an innovative model based on machine learning address these challenges. By leveraging data-driven approaches, this work establishes a workflow media. The study employs hybrid CNN-BiLSTM-Attention model, incorporating convolutional neural networks (CNN), bidirectional long short-term memory (BiLSTM), attention mechanism predict using pore-throat parameters. dataset, generated Quartet Structure Generation Set method pore network models, consists of 600 randomly created samples. Key finding include: (1) the proposed outperforms traditional models (MLP, CNN, CNN-BiLSTM), with RMSE, MAE, R2 values 0.0076, 0.0058, 0.97, respectively; (2) most influential factors affecting are mean radius, throat porosity; (3) successfully predicts oil reservoir samples, closely matching results. offers highly efficient accurate prediction, particularly suited unconventional providing potential applications evaluation enhanced recovery strategies.

Language: Английский

---

Prediction model for grouting volume using borehole image features and explainable artificial intelligence

Construction and Building Materials, Journal Year: 2025, Volume and Issue: 470, P. 140626 - 140626

Published: Feb. 28, 2025

Language: Английский

---

Analysis of MHD Viscous Fluid Flow under the Influence of Viscous Dissipation Force Over Vertically Moving Plate with Innovative Constant Proportional Caputo Derivative

Partial Differential Equations in Applied Mathematics, Journal Year: 2025, Volume and Issue: unknown, P. 101163 - 101163

Published: March 1, 2025

Language: Английский

---

Introducing a New Genetic Operator Based on Differential Evolution for the Effective Training of Neural Networks

Computers, Journal Year: 2025, Volume and Issue: 14(4), P. 125 - 125

Published: March 28, 2025

Artificial neural networks are widely established models used to solve a variety of real-world problems in the fields physics, chemistry, etc. These machine learning contain series parameters that must be appropriately tuned by various optimization techniques order effectively address they face. Genetic algorithms have been many cases recent literature train artificial networks, and modifications made enhance this procedure. In article, incorporation novel genetic operator into is proposed networks. The new based on differential evolution technique, it periodically applied randomly selected chromosomes from population. Furthermore, determine promising range values for network, an additional algorithm executed before execution basic algorithm. modified classification regression datasets, results reported compared with those other methods

Language: Английский

---

Special collection on “Recent Advancements in Data-Centric Geotechnics” for Computers and Geotechnics

Computers and Geotechnics, Journal Year: 2024, Volume and Issue: 171, P. 106415 - 106415

Published: May 14, 2024

Language: Английский

---

Prediction of hydraulic conductivity of sand with multivariate-index properties using optimal machine-learning-based regression models

Environmental Earth Sciences, Journal Year: 2024, Volume and Issue: 83(18)

Published: Sept. 1, 2024

Language: Английский

---