Revolutionizing air quality forecasting: Fusion of state-of-the-art deep learning models for precise classification

Urban Climate, Год журнала: 2025, Номер 59, С. 102308 - 102308

Опубликована: Янв. 28, 2025

Язык: Английский

---

Artificial Intelligence in Supply Chain Management: Optimizing Logistics and Efficiency

Опубликована: Янв. 1, 2025

Язык: Английский

---

Predicting E-Consumer Preferences for Collection and Delivery Points Using Deep Learning

IGI Global eBooks, Год журнала: 2025, Номер unknown, С. 395 - 412

Опубликована: Фев. 25, 2025

Collection and delivery points (CDPs) are crucial for addressing last-mile logistics challenges, particularly with the rapid growth of e-commerce. This study examines factors influencing e-consumers' preferences CDPs in Morocco, using deep learning (DL) algorithms to develop a predictive model. A dataset 3,000 responses, including age, income, accessibility CDPs, was used train models. Convolutional neural networks (CNN), long short-term memory (LSTM), FT-transformer were employed classify preferences. Hyperparameter tuning optimized these models, achieving 92% accuracy, outperforming CNN LSTM. The findings offer actionable insights providers better meet consumer needs, enhancing satisfaction operational efficiency. chapter also discusses study's limitations suggests future research directions logistics.

Язык: Английский

---

SwinConvNeXt: a fused deep learning architecture for Real-time garbage image classification

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Март 7, 2025

Abstract Waste management handles all kinds of waste, including household, industrial, municipal, organic, biomedical, biological, and radioactive wastes. People still face challenges in proper disposal methods for different types landfill-bound items, recyclable materials, biodegradable waste. Inadequate waste poses a significant multifaceted global challenge. The conventional method segregating is time-consuming ineffective that wastes human power money. To address this issue real time, sophisticated sustainable systems need to be implemented. latest advancements computer vision deep learning offer efficient solutions effective recycling management. Existing models exhibited various limitations, such as detection accuracy computational inefficiency, particularly when dealing with objects varying sizes exhibiting high degrees visual similarity. These limitations generate effectively capturing representing the nuanced features visually similar objects. problem, we proposed stacking an enhanced Swin Transformer, improved ConvNeXt, spatial attention mechanism. transformers incorporate two key components- hierarchical feature extraction shifting window mechanism extract from garbage images effectively. extracts most important regions identify In contrast, captures long-range dependencies within image garbage. ConvNext block optimized parameterization local image. This capability enables model discern fine-grained details individual particles, shape, texture, subtle variations color appearance, leading more accurate classification results. When evaluated performance using publicly available Garbage Classification dataset, it attained 98.97% accuracy, 98.42% Precision, 98.61% Recall. Due its lightweight low time power, surpasses existing state-of-the-art models.

Язык: Английский

---

A Multi-Objective Game Theory Model for Sustainable Profitability in the Tourism Supply Chain: Integrating Human Resource Management and Artificial Neural Networks

Systems and Soft Computing, Год журнала: 2025, Номер unknown, С. 200217 - 200217

Опубликована: Март 1, 2025

Язык: Английский

---

Assessing applications of emerging technologies in a data-driven era of sustainability and challenges

Technological Forecasting and Social Change, Год журнала: 2025, Номер unknown, С. 124110 - 124110

Опубликована: Март 1, 2025

Язык: Английский

---