Urban air pollution is a pressing global issue driven by factors such as swift urbanization, population expansion, and heightened industrial activities. To address this challenge, the integration of Machine Learning (ML) into smart cities presents a promising avenue. Our article offers comprehensive insights into recent advancements in air quality research, employing the PRISMA method as a cornerstone for the reviewing process, while simultaneously exploring the application of frequently employed ML methodologies. Focusing on supervised learning algorithms, the study meticulously analyzes air quality data, elucidating their unique benefits and challenges. These frequently employed ML techniques, including LSTM (Long Short-Term Memory), RF (Random Forest), ANN (Artificial Neural Networks), and SVR (Support Vector Regression), are instrumental in our quest for cleaner, healthier urban environments. By accurately predicting key pollutants such as particulate matter (PM), nitrogen oxides (NO_x), carbon monoxide (CO), and ozone (O₃), these methods offer tangible solutions for society. They enable informed decision-making for urban planners and policymakers, leading to proactive, sustainable strategies to combat urban air pollution. As a result, the well-being and health of urban populations are significantly improved. In this revised abstract, the importance of frequently employed ML methods in the context of air quality is explicitly emphasized, underlining their role in improving urban environments and enhancing the well-being of urban populations.

In the context of Supply Chain Management 4.0, costumers’ demand forecasting has a crucial role within an industry in order to maintain the balance between the demand and supply, thus improve the decision making. Throughout the Supply Chain (SC), a large amount of data is generated. Artificial Intelligence (AI) can consume this data in order to allow each actor in the SC to gain in performance but also to better know and understand the customer. This study is carried out in order to improve the performance of the demand forecasting system of the SC based on Deep Learning methods, including Auto-Regressive Integrated Moving Average (ARIMA) and Long Short-Term Memory (LSTM) using historical transaction record of a company. The experimental results enable to select the most efficient method that could provide better accuracy than the tested methods.

Breast cancer is a significant medical and social issue that draws attention from the global scientific community. The classification of ultrasound images of breast cancer plays a crucial role in computer-aided diagnostic systems. In this paper, we present a deep learning technique for the diagnosis of breast cancer using ultrasound images. The proposed system leverages Transfer Learning, a machine learning approach that allows the reuse of pre-trained models for new tasks. Four pre-trained models, namely VGG16, VGG19, MobileNetV2, and ResNet50V2, were considered for building the system. Our study uses a public dataset that contains 9016 ultrasound images related to benign and malignant breast cancers. Results indicate that the VGG19 network is the most accurate model in classifying breast cancer as benign or malignant. The proposed system showed remarkable results, with an accuracy rate of 98.44% as obtained from the VGG19 model. This study emphasizes the ability of Transfer Learning to improve the precision of breast tumor classification based on Ultrasound Images.