Bouhenni Sarra
Ecole Nationale Superieur D'informatique, Systèmes Informatiques, Graduate Student
- PhD student at the national higher school of computer science of Algiers ESI, interested in subjects related to Artificial Intelligence, Machine Learning, Data Mining and Big Data.edit
—L'acheminement des données multimédia nécessite un protocole de transport qui prend en compte la contrainte du temps réel qu'impose le visionnage direct du contenu, ce qui n'est pas garanti avec les protocoles TCP/UDP. Le protocole... more
—L'acheminement des données multimédia nécessite un protocole de transport qui prend en compte la contrainte du temps réel qu'impose le visionnage direct du contenu, ce qui n'est pas garanti avec les protocoles TCP/UDP. Le protocole Real-Time Transport Protocol (RTP/RTCP) remédie aux limites des protocoles TCP/UDP en permettant une gestion dynamique du trafic de transmission et le véhicule des informations sur la qualité de services entre la source et la destination. Dans cet article, nous proposons une nouvelle stratégie de régula-tion de débit et détection de congestion du protocole de transport MyRTP, basé sur RTP/RTCP. Ensuite, nous exposons les résultats obtenus sous le simulateur NS2, dans deux environnements filaire et mobile avec les protocoles de routages AODV et DSDV.
Research Interests:
Genetic algorithms are suitable for optimization problems that possess multiple local optima. They can provide excellent approximative solutions to a wide range of problems that either takes an infite timeto solve through an exhaustive... more
Genetic algorithms are suitable for optimization problems that possess multiple local optima. They can provide excellent approximative solutions to a wide range of problems that either takes an infite timeto solve through an exhaustive search or takes requires a huge amount of computational time. In this article, we are going to describe the implementation of genetic algorithms on the unbounded knapsack problem (UKP), which is an optimazation problem NP-hard. Most search papers had dealt only with the simplest version of knapsack problem which is the 0/1 knapsack problem. They provided algorithms that can find near-optimal solutions. However, there is not enough studies that duilt with the version of the unbounded knapsack problem (UKP) with multiple constraints, which is a strongly NP-hard combinatorial optimization problem occurring in many different applications. The goal of this article is to present an to approach to solve this problem with a sophisticated genetic algorithm.
Research Interests:
—L'acheminement des données multimédia nécessite un protocole de transport qui prend en compte la contrainte du temps réel qu'impose le visionnage direct du contenu, ce qui n'est pas garanti avec les protocoles TCP/UDP. Le protocole... more
—L'acheminement des données multimédia nécessite un protocole de transport qui prend en compte la contrainte du temps réel qu'impose le visionnage direct du contenu, ce qui n'est pas garanti avec les protocoles TCP/UDP. Le protocole Real-Time Transport Protocol (RTP/RTCP) remédie aux limites des protocoles TCP/UDP en permettant une gestion dynamique du trafic de transmission et le véhicule des informations sur la qualité de services entre la source et la destination. Dans cet article, nous proposons une nouvelle stratégie de régula-tion de débit et détection de congestion du protocole de transport MyRTP, basé sur RTP/RTCP. Ensuite, nous exposons les résultats obtenus sous le simulateur NS2, dans deux environnements filaire et mobile avec les protocoles de routages AODV et DSDV.
Research Interests:
Since man invented agriculture, plant disease epidemics have been a major challenge for crop growers. Their rapid spread can cause huge crop losses and thus threating the food and nutritional security of millions of people at the same... more
Since man invented agriculture, plant disease epidemics have been a major challenge for crop
growers. Their rapid spread can cause huge crop losses and thus threating the food and nutritional
security of millions of people at the same time.
In this report, we discuss the role of machine learning techniques in the control and monitoring
of plant diseases. In fact, thanks to classification algorithms, it is possible to build a predictive
model of a plant disease by examining the historical data of climatic conditions that have already
promoted its occurrence. Therefore, this will help detect the danger sufficiently in advance to
inform farmers of its imminence so they can take necessary precautions.
The objective of our study is to determine the evaluation criteria of classification algorithms
performance, to compare between them and select the best one for plant disease forecasting.
Therefore, through an in-depth research on classification algorithms and available comparison
methods, we have shown that the best way to select an algorithm is to try them all on the same
data set. The steps of this approach are illustrated by an example of a comparative test made with
Weka tool on a set of apple scab observations.
In parallel, throughout the literature we find that the systems using machine learning to predict
the occurrence of plant diseases are used so often in some parts of the globe, particularly United
States. This is because of the collaboration between universities in different fields and the
exchange of the necessary knowledge for the implementation of such systems. Unlike Algeria,
where studies on plant epidemics are only limited to their chemical and biological aspects.
growers. Their rapid spread can cause huge crop losses and thus threating the food and nutritional
security of millions of people at the same time.
In this report, we discuss the role of machine learning techniques in the control and monitoring
of plant diseases. In fact, thanks to classification algorithms, it is possible to build a predictive
model of a plant disease by examining the historical data of climatic conditions that have already
promoted its occurrence. Therefore, this will help detect the danger sufficiently in advance to
inform farmers of its imminence so they can take necessary precautions.
The objective of our study is to determine the evaluation criteria of classification algorithms
performance, to compare between them and select the best one for plant disease forecasting.
Therefore, through an in-depth research on classification algorithms and available comparison
methods, we have shown that the best way to select an algorithm is to try them all on the same
data set. The steps of this approach are illustrated by an example of a comparative test made with
Weka tool on a set of apple scab observations.
In parallel, throughout the literature we find that the systems using machine learning to predict
the occurrence of plant diseases are used so often in some parts of the globe, particularly United
States. This is because of the collaboration between universities in different fields and the
exchange of the necessary knowledge for the implementation of such systems. Unlike Algeria,
where studies on plant epidemics are only limited to their chemical and biological aspects.