A partir de cette page vous pouvez :
Retourner au premier Ă©cran avec les recherches... |
Unité des masters
93 résultat(s)
Affiner la recherche
Affiner la recherche
Segmentation de la texture dans les images en couleur / Oussama JAAFAR
Titre : Segmentation de la texture dans les images en couleur Type de document : projet fin Ă©tudes Auteurs : Oussama JAAFAR, Auteur Langues : Français (fre) CatĂ©gories : BIG DATA Index. dĂ©cimale : mast 249/19 RĂ©sumĂ© : La segmentation des images est un problème crucial pour l’analyse et la compréhension des images.
Parmi les approches existantes pour résoudre ce problème, la classification non supervisée est fréquemment
employée lors d’une première étape pour réaliser un partitionnement de l’espace des intensités
des pixels (qu’il s’agisse de niveaux de gris, de couleurs ou de réponses spectrales). Dans notre travail,
nous présentons une approche de la classification non supervisée permettant d’effectuer la segmentation
des images sans faire appel à des techniques supplémentaires. Plus précisément, nous élaborons
une méthode itérative de type k-means où les données à partitionner sont les pixels eux mêmes et les
histogrammes. Nous illustrons finalement le potentiel de l’approche proposée par quelques résultats
préliminaires de segmentation sur des images artificielles.
Segmentation de la texture dans les images en couleur [projet fin Ă©tudes] / Oussama JAAFAR, Auteur . - [s.d.].
Langues : Français (fre)
CatĂ©gories : BIG DATA Index. dĂ©cimale : mast 249/19 RĂ©sumĂ© : La segmentation des images est un problème crucial pour l’analyse et la compréhension des images.
Parmi les approches existantes pour résoudre ce problème, la classification non supervisée est fréquemment
employée lors d’une première étape pour réaliser un partitionnement de l’espace des intensités
des pixels (qu’il s’agisse de niveaux de gris, de couleurs ou de réponses spectrales). Dans notre travail,
nous présentons une approche de la classification non supervisée permettant d’effectuer la segmentation
des images sans faire appel à des techniques supplémentaires. Plus précisément, nous élaborons
une méthode itérative de type k-means où les données à partitionner sont les pixels eux mêmes et les
histogrammes. Nous illustrons finalement le potentiel de l’approche proposée par quelques résultats
préliminaires de segmentation sur des images artificielles.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 249/19 mast 249/19 OUS Texte imprimé Unité des masters Mast/19 Disponible Skateboard Realization & Modeling / EL ACHAACHI HICHAME
Titre : Skateboard Realization & Modeling Type de document : projet fin études Auteurs : EL ACHAACHI HICHAME, Auteur Langues : Français (fre) Catégories : Systèmes Embarqués pour l'Automobile Index. décimale : mast 291/19 Résumé : An electric motor skateboard was designed and built to test and develop an innovative propulsion system and engine controller. The prototype electric skateboard can reach a speed of more than 35 km/h and a range of more than 30 km with a single battery charge.
The prototype weighs between 7 and 9 kg and can easily be carried by the user. This mode of transport has potential uses in leisure, motor sports (races), short commutes, and most notably, in ‘the last mile’ of public transport – getting to and from a train station, bus stop, etc. to the user’s final destination..
Typical electric powered skateboards use external motors(s) requiring a power transmission assembly to drive the wheels. The hub motor design places the motor(s) inside the skateboard wheels and drives the wheels directly. This removes the need for power transmission assemblies therefore reductions in size, weight, cost, audible noise, and maintenance are realized. The hub motor built for this prototype has proven to be a highly feasible option over typical drive systems and further improvements to the design are discussed in this report.
To meet our customer’s expectations, it is necessary to understand the requirements and realizations in the form of a prototype in order to be able to analyze the results and modify the errors to have a finished product. This is why there is the modelling for the prototype realize in the form of a block under Matlab/Simulink that will help us to obtain optimal results with a low prices in a short time.Skateboard Realization & Modeling [projet fin Ă©tudes] / EL ACHAACHI HICHAME, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : Systèmes Embarqués pour l'Automobile Index. décimale : mast 291/19 Résumé : An electric motor skateboard was designed and built to test and develop an innovative propulsion system and engine controller. The prototype electric skateboard can reach a speed of more than 35 km/h and a range of more than 30 km with a single battery charge.
The prototype weighs between 7 and 9 kg and can easily be carried by the user. This mode of transport has potential uses in leisure, motor sports (races), short commutes, and most notably, in ‘the last mile’ of public transport – getting to and from a train station, bus stop, etc. to the user’s final destination..
Typical electric powered skateboards use external motors(s) requiring a power transmission assembly to drive the wheels. The hub motor design places the motor(s) inside the skateboard wheels and drives the wheels directly. This removes the need for power transmission assemblies therefore reductions in size, weight, cost, audible noise, and maintenance are realized. The hub motor built for this prototype has proven to be a highly feasible option over typical drive systems and further improvements to the design are discussed in this report.
To meet our customer’s expectations, it is necessary to understand the requirements and realizations in the form of a prototype in order to be able to analyze the results and modify the errors to have a finished product. This is why there is the modelling for the prototype realize in the form of a block under Matlab/Simulink that will help us to obtain optimal results with a low prices in a short time.RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 291/19 mast 291/19 ELA Texte imprimé Unité des masters Mast/19 Disponible Smart Agriculture / Oumaima EL GADI
Titre : Smart Agriculture Type de document : projet fin études Auteurs : Oumaima EL GADI, Auteur Langues : Français (fre) Catégories : BIG DATA Mots-clés : Irrigation management system, Evolution detection, Neural network, Internet of
ThingsIndex. décimale : mast 267/19 Résumé : This document describes the proposed system for remote control of the plant, i.e.
automating irrigation and taking the decision if the plant is suitable for the climate
or if its position needs to be changed.
Our system estimates the irrigation needs of a plantation, based on soil measurements
and climate variables collected by sensors connected with the Raspberry.
To estimate the plant’s needs, we propose to use a deep learning technique, more
precisely the Multilayer Perceptron (MLP). For prediction the algorithm is implemented,
the prediction function works on any database. By detecting if the plant
needs water, we start the pump and we will record this event.
The monitoring of the plant is synchronized periodically, we pray in consideration
of the autumn season. We used the R-CNN Mask for object detection and instance
segmentation to delimit the plant in the image and then we use the histogram
to calculate its size. The calculation is applied to the current image of the plant and
the previous one, then we compare the results if the plant evolves we display that
no need to change its environment. We have also implemented a function as a complement
to this one, which aims to detect if the fruit is ripe or not. This function is
dedicated to red fruits.
Smart Agriculture [projet fin Ă©tudes] / Oumaima EL GADI, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Mots-clés : Irrigation management system, Evolution detection, Neural network, Internet of
ThingsIndex. décimale : mast 267/19 Résumé : This document describes the proposed system for remote control of the plant, i.e.
automating irrigation and taking the decision if the plant is suitable for the climate
or if its position needs to be changed.
Our system estimates the irrigation needs of a plantation, based on soil measurements
and climate variables collected by sensors connected with the Raspberry.
To estimate the plant’s needs, we propose to use a deep learning technique, more
precisely the Multilayer Perceptron (MLP). For prediction the algorithm is implemented,
the prediction function works on any database. By detecting if the plant
needs water, we start the pump and we will record this event.
The monitoring of the plant is synchronized periodically, we pray in consideration
of the autumn season. We used the R-CNN Mask for object detection and instance
segmentation to delimit the plant in the image and then we use the histogram
to calculate its size. The calculation is applied to the current image of the plant and
the previous one, then we compare the results if the plant evolves we display that
no need to change its environment. We have also implemented a function as a complement
to this one, which aims to detect if the fruit is ripe or not. This function is
dedicated to red fruits.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 267/19 mast 267/19 OUM Texte imprimé Unité des masters Mast/19 Disponible Smart Company, Entreprise Level Problem Solving Through AI, Skill Requirement management: CV indexing and analysis / Alaoui Ismaili Soukaina
Titre : Smart Company, Entreprise Level Problem Solving Through AI, Skill Requirement management: CV indexing and analysis Type de document : projet fin études Auteurs : Alaoui Ismaili Soukaina, Auteur Langues : Français (fre) Catégories : BIG DATA Mots-clés : Natural Language Processing, CV Parsing, Machine learning, Enterprise
Resource Planning, Information Retrieval, Regular Expression, Electronic Document
Management, Named Entity RecognitionIndex. décimale : mast 268/19 Résumé : Smart business, Big data or articial intelligence are all terminologies that refer to
the application of modern and powerful technologies to process business-specic data.
This treatment is done automatically and far exceeds the eciency and accuracy achieved
by a human brain.
From the production chain to customer relations, logistics and human resources, arti-
cial intelligence invests companies. Combined with other technologies, the AI provides
a 360 ° view of customers, delivers predictive analytics, or provides contextualized answers
by drawing on all the data, structured or not. Today many leaders are attracted
to AI and want to know how this technology can help them transform their industry,
automate some of their processes, or help them increase their productivity and relevance,
because "If a recruiter can process 100 CV in one hour, articial intelligence can
treat a million in a few seconds with a much more rigorous, illustrates Jérémy Lamri,
founder of LAB RH, as long as the criteria are well determined."
For this purpose, rst, a presentation of the context of the project and the problematic
in question is necessary in the rst chapter. Second, the second chapter includes the
state of knowledge, the various achievements and work in progress in this area, that
is, articial intelligence in business applied to several areas such as human resources
that uses AI in several applications among them we nd the CV analysis which will
be more detailed in the following chapter explaining the method proposed with the
dierent technologies and algorithms used while indicating the steps followed. And
nally, in the last chapter, presented the objectives and results achieved, presenting
the proposed perspectives for the continuation of the project in the future.
Smart Company, Entreprise Level Problem Solving Through AI, Skill Requirement management: CV indexing and analysis [projet fin Ă©tudes] / Alaoui Ismaili Soukaina, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Mots-clés : Natural Language Processing, CV Parsing, Machine learning, Enterprise
Resource Planning, Information Retrieval, Regular Expression, Electronic Document
Management, Named Entity RecognitionIndex. décimale : mast 268/19 Résumé : Smart business, Big data or articial intelligence are all terminologies that refer to
the application of modern and powerful technologies to process business-specic data.
This treatment is done automatically and far exceeds the eciency and accuracy achieved
by a human brain.
From the production chain to customer relations, logistics and human resources, arti-
cial intelligence invests companies. Combined with other technologies, the AI provides
a 360 ° view of customers, delivers predictive analytics, or provides contextualized answers
by drawing on all the data, structured or not. Today many leaders are attracted
to AI and want to know how this technology can help them transform their industry,
automate some of their processes, or help them increase their productivity and relevance,
because "If a recruiter can process 100 CV in one hour, articial intelligence can
treat a million in a few seconds with a much more rigorous, illustrates Jérémy Lamri,
founder of LAB RH, as long as the criteria are well determined."
For this purpose, rst, a presentation of the context of the project and the problematic
in question is necessary in the rst chapter. Second, the second chapter includes the
state of knowledge, the various achievements and work in progress in this area, that
is, articial intelligence in business applied to several areas such as human resources
that uses AI in several applications among them we nd the CV analysis which will
be more detailed in the following chapter explaining the method proposed with the
dierent technologies and algorithms used while indicating the steps followed. And
nally, in the last chapter, presented the objectives and results achieved, presenting
the proposed perspectives for the continuation of the project in the future.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 268/19 mast 268/19 ALA Texte imprimé Unité des masters Mast/19 Disponible SMART URBAIN MOBILITY : SMART TRAFFIC LIGHT WITH Q -LEARNI / Imrane CHEMSEDDINE IDRISSI
Titre : SMART URBAIN MOBILITY : SMART TRAFFIC LIGHT WITH Q -LEARNI Type de document : projet fin études Auteurs : Imrane CHEMSEDDINE IDRISSI, Auteur Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 241/19 SMART URBAIN MOBILITY : SMART TRAFFIC LIGHT WITH Q -LEARNI [projet fin études] / Imrane CHEMSEDDINE IDRISSI, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Index. décimale : mast 241/19 Réservation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 241/19 mast 241/19 IMR Texte imprimé Unité des masters Mast/19 Disponible State of the Art on Deep Learning Video-Based Face Recognition and Tracking Systems / LISSAOUI El Yazid
Titre : State of the Art on Deep Learning Video-Based Face Recognition and Tracking Systems Type de document : projet fin études Auteurs : LISSAOUI El Yazid, Auteur Langues : Français (fre) Catégories : BIG DATA Mots-clés : Computer vision, video-based face recognition, face tracking, Kalman Filter, Particle Filter, face recognition system, triplet loss, CNN. Index. décimale : mast 262/19 Résumé : Nowadays, automatic face recognition systems are a prominent subject of research in the field of computer vision for their multiple and increasing applications; they gained more attention after the introduction of deep learning models that proved to be much more versatile and performant in the form of encoding-oriented CNNs. Nonetheless, video-based face recognition and tracking systems revealed to be much more challenging than still image operations; unconstrained environments, large volume of data and the spatial-temporality nature of videos push toward the need of developing robust systems that can overcome the mentioned challenges and aim, potentially, for real-time accurate performance. In that regard, research centered its attention toward this problematic and up until recently, dozens of papers have been published that deal with video-based recognition specifically.
This thesis presents the recent state-of-the-art deep learning techniques and models for video-based face recognition as well as methods for face tracking. The motivation behind this work is to find the potential building blocks to construct an automatic and autonomous system for video-based recognition for possibly real time applications. We will explore the Triplet Loss model and autoencoders for face recognition and review the most recent architectures used to achieve high performance. Moreover, the tracking problem will also be discussed through Kalman Filter and Particle filter and discuss their compatibility with the previously mentioned recognition.
State of the Art on Deep Learning Video-Based Face Recognition and Tracking Systems [projet fin Ă©tudes] / LISSAOUI El Yazid, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Mots-clés : Computer vision, video-based face recognition, face tracking, Kalman Filter, Particle Filter, face recognition system, triplet loss, CNN. Index. décimale : mast 262/19 Résumé : Nowadays, automatic face recognition systems are a prominent subject of research in the field of computer vision for their multiple and increasing applications; they gained more attention after the introduction of deep learning models that proved to be much more versatile and performant in the form of encoding-oriented CNNs. Nonetheless, video-based face recognition and tracking systems revealed to be much more challenging than still image operations; unconstrained environments, large volume of data and the spatial-temporality nature of videos push toward the need of developing robust systems that can overcome the mentioned challenges and aim, potentially, for real-time accurate performance. In that regard, research centered its attention toward this problematic and up until recently, dozens of papers have been published that deal with video-based recognition specifically.
This thesis presents the recent state-of-the-art deep learning techniques and models for video-based face recognition as well as methods for face tracking. The motivation behind this work is to find the potential building blocks to construct an automatic and autonomous system for video-based recognition for possibly real time applications. We will explore the Triplet Loss model and autoencoders for face recognition and review the most recent architectures used to achieve high performance. Moreover, the tracking problem will also be discussed through Kalman Filter and Particle filter and discuss their compatibility with the previously mentioned recognition.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 262/19 mast 262/19 LIS Texte imprimé Unité des masters Mast/19 Disponible Strong Tail Recommendation System / Ilias Azizi
Titre : Strong Tail Recommendation System Type de document : projet fin études Auteurs : Ilias Azizi, Auteur Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 238/19 Résumé : With the advent of the web and technological developments, among others, the mass
of data to be exploited or analyzed has become very large. As a result, it has become
difficult to know what data to look for and where to find them. Computer techniques
have been developed to ease this research as well as the extraction of relevant information.
The one we focus on in this work is the recommendation system. This is to
guide the user in their exploration of the data so that it finds relevant information.
We propose a novel recommendation system, based on the proposed Strong tail similarity,
Graph embedding, community detection. In our work, using user-item’s bipartite
graph and strong tail similarity, we first construct the strong tail item graph, we apply
Node2vec algorithm for graph embedding to represent items in a continuous space Rd,
and based on the strong tail item graph, we split items into communities using Louvain
community detection algorithm .We construct the embedding space for users and
communities from items embedding space, and we aggregate the latent vector of each
entity with the corresponding side information to enhance the representation of users
and items in the embedded space. we train our Neural Network model to predict rating
using users and items latent vectors with respect to the rating user-item’s matrix,
the model is used in the framework online process to help in generating relevant rank
recommendation list of items. Focusing on the case of movies recommendation, extensive
experiments on a real-world dataset demonstrate the effectiveness of the proposed
framework.Strong Tail Recommendation System [projet fin Ă©tudes] / Ilias Azizi, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Index. décimale : mast 238/19 Résumé : With the advent of the web and technological developments, among others, the mass
of data to be exploited or analyzed has become very large. As a result, it has become
difficult to know what data to look for and where to find them. Computer techniques
have been developed to ease this research as well as the extraction of relevant information.
The one we focus on in this work is the recommendation system. This is to
guide the user in their exploration of the data so that it finds relevant information.
We propose a novel recommendation system, based on the proposed Strong tail similarity,
Graph embedding, community detection. In our work, using user-item’s bipartite
graph and strong tail similarity, we first construct the strong tail item graph, we apply
Node2vec algorithm for graph embedding to represent items in a continuous space Rd,
and based on the strong tail item graph, we split items into communities using Louvain
community detection algorithm .We construct the embedding space for users and
communities from items embedding space, and we aggregate the latent vector of each
entity with the corresponding side information to enhance the representation of users
and items in the embedded space. we train our Neural Network model to predict rating
using users and items latent vectors with respect to the rating user-item’s matrix,
the model is used in the framework online process to help in generating relevant rank
recommendation list of items. Focusing on the case of movies recommendation, extensive
experiments on a real-world dataset demonstrate the effectiveness of the proposed
framework.RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 238/19 mast 238/19 ILI Texte imprimé Unité des masters Mast/19 Disponible Swarm AUVs: Target localization and tracking / Rabab Boulouchgour
Titre : Swarm AUVs: Target localization and tracking Type de document : projet fin études Auteurs : Rabab Boulouchgour, Auteur Langues : Français (fre) Catégories : BIG DATA Mots-clés : AUV, Swarm AUVs, Target Tracking, Deep learning Index. décimale : mast 248/19 Résumé : The underwater represents a very challenging environment to explore, the development
of autonomous underwater vehicles allowed accomplishing missions difficult if
not impossible for human divers. However, AUVs are met with other challenges such
as the mission time limited by batteries’ duration, thus the focus is now on developing
swarm AUVs who cooperate to accomplish tasks more efficiently.
The efficiency of a swarm navigation is constrained by the communication link and the
knowledge of the AUVs states, in this work we focus on the case of the master-slave architecture,
and we address the problem of the localization and tracking. The proposed
method is based on sparsely-allocated OFDM as a communication strategy and target
labeling protocol and LMHT-DeepMTT algorithm based on modified Hough transform
supported by DeepMTT a deep learning maneuvering tracking algorithm. Through
simulation, we evaluate the performance of the approach and find that it allows to
track the targets efficiently even in presence of important noise.
Swarm AUVs: Target localization and tracking [projet fin Ă©tudes] / Rabab Boulouchgour, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Mots-clés : AUV, Swarm AUVs, Target Tracking, Deep learning Index. décimale : mast 248/19 Résumé : The underwater represents a very challenging environment to explore, the development
of autonomous underwater vehicles allowed accomplishing missions difficult if
not impossible for human divers. However, AUVs are met with other challenges such
as the mission time limited by batteries’ duration, thus the focus is now on developing
swarm AUVs who cooperate to accomplish tasks more efficiently.
The efficiency of a swarm navigation is constrained by the communication link and the
knowledge of the AUVs states, in this work we focus on the case of the master-slave architecture,
and we address the problem of the localization and tracking. The proposed
method is based on sparsely-allocated OFDM as a communication strategy and target
labeling protocol and LMHT-DeepMTT algorithm based on modified Hough transform
supported by DeepMTT a deep learning maneuvering tracking algorithm. Through
simulation, we evaluate the performance of the approach and find that it allows to
track the targets efficiently even in presence of important noise.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 248/19 mast 248/19 RAB Texte imprimé Unité des masters Mast/19 Disponible Système d'authentification centralisĂ©e SSO (Single Sign-On : une seule authentification Pour plusieurs applications) / Hamza BARY / Mustafa EL AZZAOUI
Titre : Système d'authentification centralisĂ©e SSO (Single Sign-On : une seule authentification Pour plusieurs applications) Type de document : projet fin Ă©tudes Auteurs : Hamza BARY / Mustafa EL AZZAOUI, Auteur Langues : Français (fre) CatĂ©gories : Internet des Objets et Services Mobiles ( IOSM ) Mots-clĂ©s : Authentification, SSO, Ldap, CAS Index. dĂ©cimale : mast 202/19 RĂ©sumĂ© : Le présent rapport est une synthèse du travail effectué dans le cadre de notre projet de fin d'études du master IOSM.
La multiplicité des systèmes et des applications qui nous utilisons chaque jour, nous exigent de mémoriser pour chaque ressource utilisé ses identifiants et son mot de passe, ces dizaines couple (Login/Password) est devenu une réelle problématique
L'objectif de ce projet est la mise en place d’un système d’authentification unique (single Sing-On ) pour les différents applications web pour l’institut INPT, Ce system permet aux utilisateurs d’accéder au différentes applications avec une seule authentification, et donc un seul login et mot de passe.
Ce mémoire est organisé en cinq parties principales et contient une description détaillée des différentes phases de la réalisation du projet. Il s'intéresse d'une part à la partie conceptuelle afin de consolider et concevoir une solution à la problématique posée, et d'autre part à la mise en place et l’intégration de la solution obtenue avec les applications
Système d'authentification centralisée SSO (Single Sign-On : une seule authentification Pour plusieurs applications) [projet fin études] / Hamza BARY / Mustafa EL AZZAOUI, Auteur . - [s.d.].
Langues : Français (fre)
CatĂ©gories : Internet des Objets et Services Mobiles ( IOSM ) Mots-clĂ©s : Authentification, SSO, Ldap, CAS Index. dĂ©cimale : mast 202/19 RĂ©sumĂ© : Le présent rapport est une synthèse du travail effectué dans le cadre de notre projet de fin d'études du master IOSM.
La multiplicité des systèmes et des applications qui nous utilisons chaque jour, nous exigent de mémoriser pour chaque ressource utilisé ses identifiants et son mot de passe, ces dizaines couple (Login/Password) est devenu une réelle problématique
L'objectif de ce projet est la mise en place d’un système d’authentification unique (single Sing-On ) pour les différents applications web pour l’institut INPT, Ce system permet aux utilisateurs d’accéder au différentes applications avec une seule authentification, et donc un seul login et mot de passe.
Ce mémoire est organisé en cinq parties principales et contient une description détaillée des différentes phases de la réalisation du projet. Il s'intéresse d'une part à la partie conceptuelle afin de consolider et concevoir une solution à la problématique posée, et d'autre part à la mise en place et l’intégration de la solution obtenue avec les applications
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 202/19 mast 202/19 HAM Texte imprimé Unité des masters Mast/19 Disponible Système de dĂ©tection de fuite de gaz / Mohamed BOUKHRIS
Titre : Système de dĂ©tection de fuite de gaz Type de document : projet fin Ă©tudes Auteurs : Mohamed BOUKHRIS, Auteur Langues : Français (fre) CatĂ©gories : Internet des Objets et Services Mobiles ( IOSM ) Mots-clĂ©s : Gaz LPG, fuite de gaz, capteur MQ-6, ESP32 Index. dĂ©cimale : mast 200/19 RĂ©sumĂ© : Le présent document constitue une synthèse de mon projet de fin d’études, effectué au sein de la société heliantha, qui interagit dans le domaine de la domotique et l’internet des objets et plus précisément dans la détection des fuites de gaz.
En effet Le gaz de pétrole liquéfié (GPL) est devenu une source très répandue Soit pour le chauffage, la production d’eau chaude sanitaire ou pour la cuisson dans les foyers. Cependant, les fuites de GPL représentent une menace sérieuse pour l'utilisateur. Pour éviter le danger associé à l'utilisation de GPL au foyer, ce système a été mis au point pour permettre la détection rapide des fuites de gaz et la notification du problème. Un module WIFI est utilisé pour envoyer des notifications à l'utilisateur dans le cas où une fuite de gaz est détectée.
Système de détection de fuite de gaz [projet fin études] / Mohamed BOUKHRIS, Auteur . - [s.d.].
Langues : Français (fre)
CatĂ©gories : Internet des Objets et Services Mobiles ( IOSM ) Mots-clĂ©s : Gaz LPG, fuite de gaz, capteur MQ-6, ESP32 Index. dĂ©cimale : mast 200/19 RĂ©sumĂ© : Le présent document constitue une synthèse de mon projet de fin d’études, effectué au sein de la société heliantha, qui interagit dans le domaine de la domotique et l’internet des objets et plus précisément dans la détection des fuites de gaz.
En effet Le gaz de pétrole liquéfié (GPL) est devenu une source très répandue Soit pour le chauffage, la production d’eau chaude sanitaire ou pour la cuisson dans les foyers. Cependant, les fuites de GPL représentent une menace sérieuse pour l'utilisateur. Pour éviter le danger associé à l'utilisation de GPL au foyer, ce système a été mis au point pour permettre la détection rapide des fuites de gaz et la notification du problème. Un module WIFI est utilisé pour envoyer des notifications à l'utilisateur dans le cas où une fuite de gaz est détectée.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 200/19 mast 200/19 MOH Texte imprimé Unité des masters Mast/19 Disponible Systèmes de recommandation dans des rĂ©seaux sociaux / Oussoulous nabila
Titre : Systèmes de recommandation dans des rĂ©seaux sociaux Type de document : projet fin Ă©tudes Auteurs : Oussoulous nabila, Auteur Langues : Français (fre) CatĂ©gories : BIG DATA Mots-clĂ©s : Filtrage collaboratif, Filtrage par contenu, Filtrage hybride, Recommandation Index. dĂ©cimale : mast 253/19 RĂ©sumĂ© : Les médias sociaux ne sont plus un outil de communication critique; il s'est cimenté comme un aspect essentiel de toute stratégie de marketing et de communication. Voilà pourquoi l'analyse des données des médias sociaux est devenu une nécessité, nous nous intéressions à la recherche d'informations sur internet, afin de déterminer les préférences des utilisateurs , à travers notre travail, nous avons étudié les approches de recommandation, l’approche à base de contenu, l’approche à base de filtrage collaboratif, le filtrage hybride, le filtrage à base de connaissance ainsi que le filtrage démographique pour mieux assister les utilisateurs dans leurs recherches, afin de présenter de bons résultats dans le domaine des systèmes de recommandation.
Ce travail vise à présenter plusieurs méthodes, et les utiliser dans un système de recommandation qui prend en compte des données extraites. Il implémente des techniques de suggestion d'amis pour une nouvelle application web.
Systèmes de recommandation dans des réseaux sociaux [projet fin études] / Oussoulous nabila, Auteur . - [s.d.].
Langues : Français (fre)
CatĂ©gories : BIG DATA Mots-clĂ©s : Filtrage collaboratif, Filtrage par contenu, Filtrage hybride, Recommandation Index. dĂ©cimale : mast 253/19 RĂ©sumĂ© : Les médias sociaux ne sont plus un outil de communication critique; il s'est cimenté comme un aspect essentiel de toute stratégie de marketing et de communication. Voilà pourquoi l'analyse des données des médias sociaux est devenu une nécessité, nous nous intéressions à la recherche d'informations sur internet, afin de déterminer les préférences des utilisateurs , à travers notre travail, nous avons étudié les approches de recommandation, l’approche à base de contenu, l’approche à base de filtrage collaboratif, le filtrage hybride, le filtrage à base de connaissance ainsi que le filtrage démographique pour mieux assister les utilisateurs dans leurs recherches, afin de présenter de bons résultats dans le domaine des systèmes de recommandation.
Ce travail vise à présenter plusieurs méthodes, et les utiliser dans un système de recommandation qui prend en compte des données extraites. Il implémente des techniques de suggestion d'amis pour une nouvelle application web.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 253/19 mast 253/19 OUS Texte imprimé Unité des masters Mast/19 Disponible Systèmes de transport intelligent / ZRIGUI Ismail
Titre : Systèmes de transport intelligent Type de document : projet fin études Auteurs : ZRIGUI Ismail, Auteur Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 233/19 Systèmes de transport intelligent [projet fin études] / ZRIGUI Ismail, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Index. décimale : mast 233/19 Réservation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 233/19 mast 233/19 ZRI Texte imprimé Unité des masters Mast/19 Disponible Les techniques d’échantillonnage dans le cadre de Big Data / MAJDALANE Siham
Titre : Les techniques d’échantillonnage dans le cadre de Big Data Type de document : projet fin études Auteurs : MAJDALANE Siham, Auteur Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 254/19 Résumé : The objective of sampling techniques is to extract a sample that has the same characteristics as the original database.
This document summarizes several ideas, on the field of Big Data and its charac- teristics, as well as the tools used for data preparation.
This report also provides a general introduction to sampling techniques, and provides a general review of the literature on the various works that have been done to test the performance and reliability of these techniques to see if the results obtained by the use of these methods can be trusted, and also discusses the contributions of these techniques to data mining, and also gives some proposals for advancing the state of the art in sampling techniques in the context of Big Data.Les techniques d’échantillonnage dans le cadre de Big Data [projet fin études] / MAJDALANE Siham, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Index. décimale : mast 254/19 Résumé : The objective of sampling techniques is to extract a sample that has the same characteristics as the original database.
This document summarizes several ideas, on the field of Big Data and its charac- teristics, as well as the tools used for data preparation.
This report also provides a general introduction to sampling techniques, and provides a general review of the literature on the various works that have been done to test the performance and reliability of these techniques to see if the results obtained by the use of these methods can be trusted, and also discusses the contributions of these techniques to data mining, and also gives some proposals for advancing the state of the art in sampling techniques in the context of Big Data.RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 254/19 mast 254/19 MAJ Texte imprimé Unité des masters Mast/19 Disponible Toward a recommender system for helping learners at risk of dropping out in MOOCs / SEKKOU Amjad
Titre : Toward a recommender system for helping learners at risk of dropping out in MOOCs Type de document : projet fin études Auteurs : SEKKOU Amjad, Auteur Langues : Français (fre) Catégories : BIG DATA Index. décimale : mast 263/19 Résumé : Recently, Massive Open Online Courses (MOOCs) have aroused a great interest in
the media. As the number of students enrolled in MOOCs increases, and given
the lack of supervision, it becomes dicult for students to get answers to their
questions through discussion forums, resulting a very high attrition rate. In this
work, we addressed the problem of unanswered questions in the discussion forums,
based on a combination of Social Network Analysis and Deep Learning. First, we
analyzed a network that presents the interaction between students, and an other that
illustrates the structure of the threads in order to obtain the lone questions. Then,
we calculated a similarity score between these questions, using our proposed model.
This last, is the key element of the semantic similarity approach between forum
questions. The results of our experience on a Stanford University MOOC course
show that our recommendation method has the potential to guide students to the
answers of their questions, and also to learners who can help them. Thus, achieve
the main goal of this master thesis: Reduce the high attrition rate in MOOCs.Toward a recommender system for helping learners at risk of dropping out in MOOCs [projet fin Ă©tudes] / SEKKOU Amjad, Auteur . - [s.d.].
Langues : Français (fre)
Catégories : BIG DATA Index. décimale : mast 263/19 Résumé : Recently, Massive Open Online Courses (MOOCs) have aroused a great interest in
the media. As the number of students enrolled in MOOCs increases, and given
the lack of supervision, it becomes dicult for students to get answers to their
questions through discussion forums, resulting a very high attrition rate. In this
work, we addressed the problem of unanswered questions in the discussion forums,
based on a combination of Social Network Analysis and Deep Learning. First, we
analyzed a network that presents the interaction between students, and an other that
illustrates the structure of the threads in order to obtain the lone questions. Then,
we calculated a similarity score between these questions, using our proposed model.
This last, is the key element of the semantic similarity approach between forum
questions. The results of our experience on a Stanford University MOOC course
show that our recommendation method has the potential to guide students to the
answers of their questions, and also to learners who can help them. Thus, achieve
the main goal of this master thesis: Reduce the high attrition rate in MOOCs.RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 263/19 mast 263/19 SEK Texte imprimé Unité des masters Mast/19 Disponible Towards a serendipitous e-learning recommendation system / Zahra QARNOUF
Titre : Towards a serendipitous e-learning recommendation system Type de document : projet fin Ă©tudes Auteurs : Zahra QARNOUF, Auteur Langues : Français (fre) CatĂ©gories : BIG DATA Mots-clĂ©s : Recommender system, Serendipity, E-learning Index. dĂ©cimale : mast 252/19 RĂ©sumĂ© : Nowadays, we are witnessing a big growth of information especially in learning technology with the advent of Internet. Thus how to recommend appropriate course to improve students’ learning outcomes has become a daunting task. With this growth, a student’s choices have grown exponentially. As a result, recommender systems have been put in place to deal with this huge amount of information, as it is difficult for students to decide which courses to choose. Therefore, we rely on the social aspect of the learning experience, to propose an algorithm that injects serendipitous items within a recommendation system. The approach should help broaden students’ horizons and provide the unexpectedness often lacking in existing e-learning platforms.
Towards a serendipitous e-learning recommendation system [projet fin Ă©tudes] / Zahra QARNOUF, Auteur . - [s.d.].
Langues : Français (fre)
CatĂ©gories : BIG DATA Mots-clĂ©s : Recommender system, Serendipity, E-learning Index. dĂ©cimale : mast 252/19 RĂ©sumĂ© : Nowadays, we are witnessing a big growth of information especially in learning technology with the advent of Internet. Thus how to recommend appropriate course to improve students’ learning outcomes has become a daunting task. With this growth, a student’s choices have grown exponentially. As a result, recommender systems have been put in place to deal with this huge amount of information, as it is difficult for students to decide which courses to choose. Therefore, we rely on the social aspect of the learning experience, to propose an algorithm that injects serendipitous items within a recommendation system. The approach should help broaden students’ horizons and provide the unexpectedness often lacking in existing e-learning platforms.
RĂ©servation
RĂ©server ce document
Exemplaires
Code barre Cote Support Localisation Section DisponibilitĂ© mast 252/19 mast 252/19 ZAH Texte imprimé Unité des masters Mast/19 Disponible