Mohammed Zouiten

Fifth-generation (5G) networks represent a paradigm shift in telecommunications, offering ultra-reliable low-latency communication, massive connectivity of devices, and unparalleled data rates. While these advantages also present significant complications surrounding energy consumption and cybersecurity, requiring new approaches to maintain operational effectiveness and network fidelity. This study proposes a new hybrid artificial intelligence (AI) framework consisting of explainable AI (XAI) for transparent resource allocation, convolutional neural networks (CNNs) for real-time anomaly detection, and recurrent neural networks (RNNs) for predictive energy optimization. Experiments and real-world case studies illustrate this framework’s scalability and efficiency by achieving improved network resource management, a detection accuracy of 99.7% for anomalies, and energy savings of up to 65%.

Groundwater is an essential source of freshwater to meet the growing water needs of agricultural, industrial, and domestic activities. the management of these resources is of crucial importance to ensure their sustainability and abundance over time and space. this major challenge requires extensive research for the prediction and simulation of groundwater level variations The new approach used in the field of hydrology and hydrogeology for the simulation and prediction of groundwater levels is based on machine learning techniques; these techniques are essential to improve the planning and management of groundwater resources. In this context, several researchers in this field have developed machine learning models capable of estimating the fluctuation of groundwater levels to protect these aquifers. This study launched word queries in several databases: Scopus, Science Direct, and IEEE; we selected 155 references and after the realization of the word clouds we find many topics: piezometer; groundwater; machine learning models, and parameters. In addition, the most recurrent machine learning models used by researchers in this field are Random Forest (RF), Decision Tree (DT), Support vector regression (SVM), artificial neural network (ANN), and SVR. But each of these researchers has based on some number of parameters as inputs; the most common are temperature; precipitation and groundwater level history. In this paper, we present a comparison of the most common machine learning models and parameters for assessing and predicting groundwater fluctuation.

Remote sensing has evolved through the appearance of several approaches. Object-based image analysis is a compelling approach to land use classification, object detection, and change detection in each environment. This paradigm is based on a critical and fundamental segmentation step. However, this step is highly dependent on the determination of the optimal parameters to be achieved. In this sense, methods have been invented to define the optimal segmentation parameters. This article presents an updated review of methods for defining optimal segmentation parameters. For this purpose, pertinent articles published in the main remote sensing journals from the emergence of the concept of object-based image analysis and segmentation to the present were used. The main aim is to provide a precise and detailed review of the different approaches previously presented. The originality of this review resides in the survey of all methods from conventional to the most recent with a discussion of these approaches. The results show that despite the advances in this field of research, most studies use the manual trial-and-error method. Conversely, state-of-the-art methods tend to determine the optimal parameter per type of geographic object and the adaptive calculation of segmentation parameters. Furthermore, the leading methods identified rely on supervised and unsupervised measures similarly, most of which use homogeneity measures. In contrast, a balance between intra- and inter-segment homogeneity and heterogeneity measures are more relevant. A distinction is made between pre-estimation and posterior parameter estimation methods.