This study empirically investigates the relationship between dataset size and classification performance in precision agriculture applications. Seven machine learning models (Decision Tree, Random Forest, Logistic Regression, SVM, Gaussian Naïve Bayes, KNN, and AdaBoost) were evaluated on seven agricultural datasets ranging from 100 to 4,000 samples. Performance was assessed using five metrics: accuracy, precision, recall, F1-score, and ROC-AUC. The methodology involved two phases: initial evaluation using complete datasets, followed by systematic analysis of subdivided datasets to examine performance variation with data volume. Statistical analysis using Pearson correlation coefficients revealed no significant correlation between dataset size and model performance (r = 0.12, p > 0.05). Results indicate that Random Forest and Decision Tree models achieved the highest average performance across datasets (88.48% and 85.37% accuracy, respectively). The findings suggest that dataset quality and problem characteristics have greater influence on classification performance than dataset size alone in precision agriculture applications.

Agriculture is confronting many challenges, from climate change to the considerable decreasing of the natural resources. In order to limit the impact of these challenges and ensure food sufficiency for the global growing population, precision agriculture is defined as one of sustainable solutions. This solution manages the whole agricultural cycle and it is based mainly on the use of information technologies. It ensures precision in the applied treatments, quantities and in time. By taking into account the nature of the agricultural field, this paper is interested in wireless technologies, more precisely, in IoT. In fact, many architectures and implementations have been proposed in this context, which can cause issues in scalability and interoperability. Thus, the principal aim of this paper is to define a conventional IoT architecture that can be used for precision agriculture.

Precision agriculture has shown its effectiveness for sustainable agriculture. It is a transversal concept and therefore covers the whole agricultural cycle. Precision agriculture principally aims to manage the agricultural process and resources. These resources are decreasing especially with the global population growth and the damaging effects of climate change. This paper focuses on the first step in implementing precision agriculture, namely the choice of the most appropriate crop. This choice has a direct impact on productivity and later, on the cost of agricultural treatments such as the fertilization process. To build the crop recommendation system, we primarily used the soil properties (pH, humidity, macronutrients NPK, rainfall and temperature). Then, we implemented several models to choose the best one. The implemented models are Fuzzy Logic, Random Forest, Gaussian Naïve Bayes, XGBoost, Logistic Regression and Artificial Neural Network. XGBoost was chosen, since it was the model that achieved the top accuracy with 99.31818%.

Several studies investigated the diagnosis of Parkinson’s disease (PD), which utilized machine learning methods such as support vector machine, neural network, Naïve Bayes and K-nearest neighbor. In addition, different ensemble methods were used such as bagging, random forest and boosting. On the other hand, different feature ranking methods have been used to reduce the data dimensionality by selecting the most important features. In this paper, the ensemble methods, random forest, XGBoost and CatBoost were used to find the most important features for predicting PD. The effect of these features with different thresholds was investigated in order to obtain the best performance for predicting PD. The results showed that CatBoost method obtained the best results.

Network virtualization is the ideal solution for the ossification internet phenomena. However, the multitude of actors involved poses significant challenges to the virtual network monitoring. For this purpose, we propose in this work a new approach for monitoring the services based on SLA established during the supply operation. An approach that aims to ensure an acceptable level of performance during all phases of the development and operation of virtual networks.

Network virtualization is an emerging concept that aims to facilitate the integration of new technologies regardless physical layer. To do this, the new concept proposes to evolve the current Internet business model. The new business model divides the Internet access provider to several actors. This actors diversity poses several challenges, especially when allocating resources and monitoring. For this, we propose in this work a new approach based on the enhanced telecom operation management business process framework. An approach that aims at ensuring an automatic supply of resources and a real-time monitoring based on the SLA defines between different actors.

The architecture of Next Generation of networks (NGN) aims to diversify the offer of operators in added value services. To do this, NGN offers a heterogeneous architecture for the services deployment. This poses significant challenges in terms of end-to-end assurance of services. For this purpose, we propose in this work the establishment of a proactive autonomous system, capable of ensuring an acceptable quality level according to Service Level Agreement (SLA) requirements. A system that is able to predict any QoS degradation due to the prediction model based on time series adapted to NGN.

Virtual networks are considered as a new concept that has emerged to support the emergence of the architecture of the Internet. A concept which leads to deploy several logical networks on the same physical medium. To do this, a new business model has emerged that divides the operator into several actors. This diversity of later poses major problems when creating virtual topologies. To this end, we propose in this work a new approach for the interfacing between these different actors. This approach is based on the eTOM framework to set up a broker-oriented business processes.

The IMS networks offer several advantages to telecom operators in terms of connectivity and deployment of new services. However, the multitude of supported access technologies poses significant challenges for QoS management. To this effect, The INQA approach (IMS Network QoS Architecture) provides innovative solutions for services supervision and monitoring, in real time. This allows identification of any QoS degradation, based in resource status. In this work we propose the integration of correction mechanisms and QoS restoration, based on eTOM processes and reports monitoring outcome of supervision operation. The work aims to set up an autonomous system for QoS correction, in IMS context

The INQA (IMS Network QoS Architecture) approach offers several scenarios for IMS (IP Multimedia Subsytem) monitoring as supervision and correction tools of services deterioration. The Platform INQA uses a set of entities and multitude technologies in monitoring. This requires deployment of considerable resources to carry out different scenarios and take more time in execution and processing. As solution, we propose in this work the integration of a new approach fuzzy logic-based. The approach optimizes costs by identifying for each situation, the appropriate scenario and related condition of service.

The new INQA approach (IMS Network QoS Architecture) for IMS network (IP Multimedia Subsystem) monitoring, provides innovative solutions for management and monitoring services, in real time. This approach aims to change the reliability of services deployed by operators in a multi technological context which raises many challenges. However, the deployment costs of management entities are significant, particularly in terms of execution time and resources load. To this end, we propose in this paper a comparison between the deployment costs via both CORBA and SOA technologies, in order to identify the most appropriate one for monitoring IMS networks.

The IMS Network QoS Architecture approach for monitoring and supervision services in the IMS offers innovative solutions for QoS management in real time. These solutions enable the integration of automatic correcting mechanisms for QoS deterioration, depending on network conditions and customer importance. However, the correction operation is costly in terms of resources and execution time. Face to this problem, the migration to an intelligent system based on reasoning based cases will enable the exploitation of recorded solutions for similar degradation cases. In this context, we propose a methodology for structuring both cases and recorded solutions in such a way as to minimize the research cost. This methodology can be held in two stages, the clustering for grouping similar cases and the classification for backup structure generation of event.

The IMS (IP Multimedia Subsystem) network as a new generation suffers of traditional Internet problems such as managing and monitoring QoS which have a direct impact on the the operator's revenue streams. Among solutions proposed, the eTOM (enhanced Telecom Operations Map) published by the TM Forum and 3GPP specifications will be sufficient to describe a system for monitoring and managing QoS. Moreover, the most of the monitoring architecture is centralized at the level of processing performance data, which make monitoring much more difficult and very slow operation and with less effective against the problems. In this paper, we propose a distributed architecture eTOM-based for IMS monitoring, and as use case the SLA verification scenario for multimedia services like VoD (Video on Demand).

The 3GPP standards for IMS (IP Multimedia System) provide access to multimedia services with robust procedures for QoS management. However its scope is limited to session initialization only, thus lacking follow-up or monitoring functionality; neither does it tackle SLA (Service Level Agreement) differentiation. A tempting approach would be to leverage the standard architecture by assurance processes based on TMForum's eTOM framework. This scenario would however necessitate a set of general business processes and would require a projection of IMS processes towards the eTOM. The work presented here follows this strategy, providing IMS services monitoring functionality based on eTOM processes able to provide SLA assessment and Service Assurance. It uses the BPEL language for orchestration and SOA elements to implement the distributed architecture.

The QoS management in IP Multimedia subsystem Networks (IMS) is a cornerstone for developing real-time services, and facility the acquisition of multitude access technologies. The 3GPP specifications proposes set of scenarios for IMS networks, which focus primarily on providing service, but without internal and external monitoring mechanisms help to correct QoS and resolve network failure. Indeed, this paper proposes a new approach for IMS networks . monitoring, which handles policing and monitoring of media-plane traffic by implementing eTOMbased business processes. The approach aims to implement a self-configurable system enable adequate monitoring and configuring depending on the business level as customer SLA, the monitoring architecture include a composite services by using web service techniques to implement a new generation of service management operations in IMS networks.

The integration of IMS networks will enable telecom operators to evolve in a manner transparent to the incessant demand for the multimedia services.However the QoS management mechanisms defined for IMS networks are considered poor in oversight and monitoring real-time services.Moreover the eTOM Framework includes the scenarios of monitoring service delivery that enable real-time tracking services being supplies.These specifications are standard and contain no specification for IMS networks.We propose in this paper a new approach to monitoring of IMS networks, and eTOM process based, the monitoring architecture is deploying by the WSOA concept.

The scope of IMS QoS management is limited to session initialization and QoS provisioning, thus lacking follow-up or monitoring functionality; neither does it tackle user differentiation. A tempting approach would be to leverage the standard 3GPP architecture by Assurance services based on TMForum's eTOM framework. This scenario would however expose a set of general business processes and would require a projection of IMS processes towards the eTOM. The work presented here follows this strategy, providing monitoring functionality to IMS services based on eTOM processes able to provide SLA assessment and Assurance services. The distributed architecture involves the BPEL language for orchestration and SOA components.