The Internet of Things (IoT) connects billions of devices across domains such as transportation, healthcare, agriculture, and energy-creating highly dynamic, distributed, and heterogeneous environments. These characteristics pose significant challenges for control, coordination, scalability, and adaptability. In response, Multi-Agent Deep Reinforcement Learning (MADRL) has emerged as a promising paradigm by combining the decision-making intelligence of reinforcement learning with the collaborative capabilities of multi-agent systems. To leverage MADRL in building intelligent, resilient, and adaptive IoT systems, this review systematically explores the application of MADRL in IoT, categorizing contributions by application domains, learning architectures, and coordination strategies. We analyze how MADRL enables scalable resource allocation, routing optimization, energy efficiency, and fault detection in complex IoT ecosystems. Furthermore, we highlight key challenges-including scalability, non-stationarity, partial observability, and communication overhead-and discuss emerging solutions such as mean-field approximation, belief-state tracking, and federated MADRL.

To address key challenges in IoT systems, including efficient resource allocation, adaptive service composition, and Quality of Service (QoS) under dynamic conditions, we develop a framework called DRL-MAS integrating multi-agent systems (MAS) and deep reinforcement learning (DRL). DRL-MAS leverages MAS's decentralized decision-making capabilities and DRL's adaptive learning strengths to ensure scalability, energy efficiency, and responsiveness in distributed IoT systems. By incorporating edge computing, DRL-MAS minimizes dependency on centralized systems, reduces latency, and optimizes energy consumption. Experimental results demonstrate the DRL-MAS's effectiveness in dynamically optimizing service composition and resource management while complying with QoS requirements.

This paper describes a proposition for constructing an automatic personalization system based on SVM (support vector machine) method. Our approach helps the learning units designers to select automatically the learning scenarios adapted to learners. In our experimentation, we have used a database that contains information about computer science engineering students of the Tlemcen university and descriptions of learning scenarios. We have implemented our SVM classifier using the open environment rdquoWekardquo. The test results showed an attractive performance. The values of the classification rate, the precision and the recall are very acceptable.

We present a new unsupervised segmentation of textural images based on integration of a texture descriptor in the formulation of active contour. The proposed texture descriptor intrinsically describes the geometry of textural regions using the shape operator defined in Beltrami framework. We use the Chernoff distance to define an active contours model which discriminates textures by maximizing the distance between the probability density functions which leads to distinguish textural objects of interest and background described by texture descriptor. We prove the existence of a solution to the new formulated active contours based segmentation model and we propose a fast and easy algorithm based on the dual formulation of the Total Variation norm. Finally, we show results on challenging images to illustrate accurate segmentations that are possible.

A new geometric active contour model based on iterative refinement of edge map stopping function obtained by iterative grey level homogenization is presented. To homogenize grey level of initial image, we proposed to couple adaptively the partial differential equation (PDE) of the anisotropic diffusion filter to that of geometric active contour model. The proposed model avoids the leakage problems and ensures that the evolving level set curves of geometric active contour model to reach more rapidly the true edges boundaries of the objects to be segmented. The robustness and precision performance of proposed model are evaluated on MR images and compared to the classical geometric active contour model (CGAC).