Fifty years of land use and land cover mapping in the United Arab Emirates: a machine learning approach using Landsat satellite data DOI Creative Commons

M. Sultan,

Salem Issa,

Basam Dahy

et al.

Frontiers in Earth Science, Journal Year: 2024, Volume and Issue: 12

Published: Dec. 11, 2024

This study analyses the spatiotemporal distribution of land use and cover (LULC) in United Arab Emirates (UAE) over past 50 years (1972–2021) using 72 multi-temporal Landsat satellite images. Three machine learning (ML) classifiers, Classification Regression Tree (CART), Support Vector Machine (SVM) Random Forest (RF), were tested, with RF finally chosen for its higher performance. Spectral, spatial, topographic, object aspect attributes extracted used as input algorithm to enhance classification accuracy. A dataset comprising 46,146 polygons representing four LULC classes was created, 80% allocated training 20% testing, ensuring robust model validation. The trained develop a that classified data into namely: built areas, vegetation, water, desert mountainous regions, producing eight thematic maps 1972, 1986, 1992, 1997, 2002, 2013, 2017, 2021. results reveal dominance their coverage gradually declining from 97% 1972 nearly 91% In contrast, areas grew less than 1% 6%, while vegetation increased 0.71% 2.85%. Water bodies have exhibited periodic fluctuations between 0.4% 0.35%. These changes are attributed extensive urbanization, agricultural expansion, forest plantation programs, reclamation, megaprojects. Accuracy assessment showed high overall accuracy, ranging 85.11% 98.4%. provides unique long-term analysis UAE years, capturing key developments 1970s oil boom through subsequent megaprojects at onset new millennium, leading reduced reliance on oil. findings underscore role geospatial technologies monitoring challenging environments, serve vital tool policymakers manage resources, urban planning, environmental conservation.

Language: Английский

Modeling to evaluate permanent gully susceptibility and dominant controlling factors analysis in the black soil region of Northeast China DOI

Wang Hong-yue,

Ruixiang Liu, Yantun Song

et al.

Soil and Tillage Research, Journal Year: 2025, Volume and Issue: 252, P. 106595 - 106595

Published: April 16, 2025

Language: Английский

Citations

0
Integrating Machine Learning and AI into IoT-Enabled Smart Parking DOI
Vesna Knights,

Olivera Petrovska,

Marija Prchkovska

et al.

Published: Jan. 1, 2025

Language: Английский

Citations

0
Machine Learning Models and Mathematical Approaches for Predictive IoT Smart Parking DOI Creative Commons
Vesna Knights,

Olivera Petrovska,

Jasmina Bunevska-Talevska

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(7), P. 2065 - 2065

Published: March 26, 2025

This paper aims to create an innovative approach improving IoT-based smart parking systems by integrating machine learning (ML) and Artificial Intelligence (AI) with mathematical approaches in order increase the accuracy of availability predictions. Three regression-based ML models, random forest, gradient boosting, LightGBM, were developed their predictive capability was compared using data collected from three locations Skopje, North Macedonia 2019 2021. The main novelty this study is based on use autoregressive modeling strategies lagged features Z-score normalization improve time series forecasts. Bayesian optimization chosen for its ability efficiently explore hyperparameter space while minimizing RMSE. able capture temporal dependencies more effectively than other resulting lower RMSE values. LightGBM model produced R2 0.9742 0.1580, making it best prediction. Furthermore, system architecture also deployed which included real-time collection sensors placed at entry exit lots individual slots. integration ML, AI, IoT technologies improves efficiency management system, reduces traffic congestion and, most importantly, offers a scalable development urban mobility solutions.

Language: Английский

Citations

0
A Solution to the Problem of Retail Credit Risk Pricing Problem Based on the Machine Learning XGBoost Algorithm DOI
Jingxuan Ma, Xin Li, J. Y. Guo

et al.

Published: Jan. 1, 2025

Language: Английский

Citations

0
Comparison of Logistic Regression, Frequency Ratio, Weight of Evidence and Shannon's Entropy Models in Erosion Susceptibility Analysis in Bingöl (Türkiye) with GIS DOI
Orhan İnik, Mustafa Utlu

Tarım Bilimleri Dergisi, Journal Year: 2025, Volume and Issue: 31(2), P. 538 - 557

Published: March 25, 2025

Soil erosion is one of the most important and critical processes occurring in Türkiye, as all parts world. It great importance to understand that occur soil continues. The aim this study determine susceptibility Çapakçur Stream basin, areas Türkiye. In study, analysis was carried out using 4 different methods Shannon Entropy (SE), Logistic Regression (LR), Frequency Ratio (FR) Weight Evidence (WoE) are effectively used today determination terms erosion, 19 conditioning factors based on these methods. Analysis Results Model performances were evaluated Receiver Operating Characteristic (ROC) Area under Curve (AUC) values a dataset consisting 840 training (70%) 360 testing (30%) points. According result AUC show regression seems perform well both (AUC= 94.7%) validating datasets (AUC=93.5%). On other hand, 93.5%) 91.4%), (AUC=92.4%) ROC similar result, but slightly lower than Regression. Additionally, shows it performs 55.7%) 56.3%). Conducting analyses methods, especially studies, will facilitate planning accuracy results obtained.

Language: Английский

Citations

0
Geospatial Analysis and Machine Learning Framework for Urban Heat Island Intensity Prediction: Natural Gradient Boosting and Deep Neural Network Regressors with Multisource Remote Sensing Data DOI Open Access
Nhat‐Duc Hoang,

Quoc-Lam Nguyen

Sustainability, Journal Year: 2025, Volume and Issue: 17(10), P. 4287 - 4287

Published: May 8, 2025

The increasing severity of the urban heat island (UHI) effect is a consequence rapid expansion and global climate change. center Da Nang, Vietnam, currently experiencing severe UHI effects combined with increasingly frequent heatwaves. This study employs advanced machine learning techniques—including natural gradient boosting deep neural network—to model spatial variation in intensity. explanatory variables include topographical features, distances to coastlines rivers, land cover types, built-up density, greenspace bareland waterbody distance wetlands. Experimental results show that models successfully explain 90% To identify primary factors influencing intensity, Shapley additive explanations are utilized. Additionally, network-based cellular automata implemented project future changes. proposed framework then employed forecast intensity Nang’s 2040. Based on prediction results, area extremely high expected increase by 3.7%. projected rise 4.6%, while medium anticipated expand 12.6%. Notably, it forecasted areas low decrease 3.9% 40.8%, respectively. findings from this can be useful assist planners establishing effective mitigation strategies for reducing impact effects.

Language: Английский

Citations

0
Traditional Cultural and Creative Product Design Methods Combining Digital Art Elements DOI Open Access

Xiaoqing Xu,

Jingxin Chen

Applied Mathematics and Nonlinear Sciences, Journal Year: 2024, Volume and Issue: 9(1)

Published: Jan. 1, 2024

Abstract As a mode of perpetuating and revitalizing traditional culture, cultural creative products have garnered widespread affection recognition from the public. In context evolving societal trends advancements in science technology, digitization has emerged as prominent trend. This study undertakes digital design these products, primarily focusing on innovative application style transfer algorithm to motifs, supplemented by their visualization through platforms such augmented reality (AR) virtual (VR). Furthermore, it facilitates intelligent consumer interaction via gesture algorithms, thereby enhancing user engagement experience. During implementation phase, this research conducts comparative analyses within products. It also employs Kano questionnaire categorize analyze needs effectively. Notably, while recall rate documented remains below 0.9, consistently achieves high precision, significantly enhances feature extraction capabilities, improves quality effects produced. Moreover, static an impressive average 98.2%. The dynamic algorithm, meanwhile, maintains 94% with processing time 3.2 seconds, balancing demands real-time accuracy systematically analyzes significance each requirement element across six dimensions, classifying customer for into four distinct categories. Additionally, delineates viable pathway integration art elements artistic setting foundation future innovations field.

Language: Английский

Citations

1
N-heterocyclic carbene coordinated single atom catalysts on C2N for enhanced nitrogen reduction DOI Open Access

Wenming Lu,

Dian Zheng,

Daifei Ye

et al.

Journal of Materials Informatics, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 28, 2024

Single-atom catalysts (SACs) with N-heterocyclic carbene (NHC) coordination provide an effective strategy for enhancing nitrogen reduction reaction (NRR) performance by modulating the electronic properties of metal active sites. In this work, we designed a novel NHC-coordinated SAC embedding transition metals (TM) into two-dimensional C2N-based nanomaterial (TM@C2N-NCM) and evaluated NRR catalytic using combination density functional theory machine learning. A multi-step screening identified eight high-performance (TM = Nb, Fe, Mn, W, V, Ta, Zr, Ti), Nb@C2N-NCM showing best (limiting potential -0.29 V). All demonstrated lower limiting values compared to their TM@graphene-NCM counterparts, revealing effectiveness C2N substrate in activity. Machine learning analysis achieved high predictive accuracy (coefficient determination 0.91; mean absolute error 0.19) final step protonation (S6), Mendeleev number (Nm), d-electron count (Nd) as key factors influencing performance. This study offers valuable insights rational design SACs highlights nanomaterials advancing electrocatalysts.

Language: Английский

Citations

1
Fifty years of land use and land cover mapping in the United Arab Emirates: a machine learning approach using Landsat satellite data DOI Creative Commons

M. Sultan,

Salem Issa,

Basam Dahy

et al.

Frontiers in Earth Science, Journal Year: 2024, Volume and Issue: 12

Published: Dec. 11, 2024

This study analyses the spatiotemporal distribution of land use and cover (LULC) in United Arab Emirates (UAE) over past 50 years (1972–2021) using 72 multi-temporal Landsat satellite images. Three machine learning (ML) classifiers, Classification Regression Tree (CART), Support Vector Machine (SVM) Random Forest (RF), were tested, with RF finally chosen for its higher performance. Spectral, spatial, topographic, object aspect attributes extracted used as input algorithm to enhance classification accuracy. A dataset comprising 46,146 polygons representing four LULC classes was created, 80% allocated training 20% testing, ensuring robust model validation. The trained develop a that classified data into namely: built areas, vegetation, water, desert mountainous regions, producing eight thematic maps 1972, 1986, 1992, 1997, 2002, 2013, 2017, 2021. results reveal dominance their coverage gradually declining from 97% 1972 nearly 91% In contrast, areas grew less than 1% 6%, while vegetation increased 0.71% 2.85%. Water bodies have exhibited periodic fluctuations between 0.4% 0.35%. These changes are attributed extensive urbanization, agricultural expansion, forest plantation programs, reclamation, megaprojects. Accuracy assessment showed high overall accuracy, ranging 85.11% 98.4%. provides unique long-term analysis UAE years, capturing key developments 1970s oil boom through subsequent megaprojects at onset new millennium, leading reduced reliance on oil. findings underscore role geospatial technologies monitoring challenging environments, serve vital tool policymakers manage resources, urban planning, environmental conservation.

Language: Английский

Citations

0