University Sétif 1 FERHAT ABBAS Faculty of Sciences
Détail de l'auteur
Auteur Semcheddine,Moussa |
Documents disponibles écrits par cet auteur
Ajouter le résultat dans votre panier Affiner la recherche
Titre : Artificial Rabbit Optimization For Tuning Deep Learning Parameters Type de document : texte imprimé Auteurs : Yasmine Tigha ; Ibtihel Boussahel ; Semcheddine,Moussa, Directeur de thèse Editeur : Setif:UFA Année de publication : 2023 Importance : 1 vol (84 f .) Format : 29 cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Optimization Artificial Rabbit Optimization Genetic Algorithms Particle
Swarm Optimization Grey Wolf Optimization Deep Learning learning rate and dropout rateIndex. décimale : 004 - Informatique Résumé : Deep learning models’ performance heavily relies on selecting appropriate hyperparameters,
such as learning rate, dropout rate , and network architecture.
However, finding the optimal values for these hyperparameters is challenging due
to the lack of fixed rules.
In this thesis, we propose using bio-inspired algorithms, including ARO, PSO,
GA, and GWO, to optimize the hyperparameters of a feed-forward neural network.
The experiments are conducted on the MNIST dataset, commonly used for image
analysis. Comparing the accuracy of models trained with these algorithms, with
and without dropout, reveals that bio-inspired algorithms improve deep learning
model accuracy. Incorporating bio-inspired algorithms in hyperparameter tuning
shows promise for optimizing deep learning models. By drawing inspiration from
nature, these algorithms enhance performance and generalization, demonstrating
their efficacy in deep learning hyperparameter tuning
Côte titre : MAI/0761 En ligne : https://drive.google.com/file/d/1WcfNow4t2G8wXoZTx2oNMTp_XM-CbHG0/view?usp=drive [...] Format de la ressource électronique : Artificial Rabbit Optimization For Tuning Deep Learning Parameters [texte imprimé] / Yasmine Tigha ; Ibtihel Boussahel ; Semcheddine,Moussa, Directeur de thèse . - [S.l.] : Setif:UFA, 2023 . - 1 vol (84 f .) ; 29 cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Optimization Artificial Rabbit Optimization Genetic Algorithms Particle
Swarm Optimization Grey Wolf Optimization Deep Learning learning rate and dropout rateIndex. décimale : 004 - Informatique Résumé : Deep learning models’ performance heavily relies on selecting appropriate hyperparameters,
such as learning rate, dropout rate , and network architecture.
However, finding the optimal values for these hyperparameters is challenging due
to the lack of fixed rules.
In this thesis, we propose using bio-inspired algorithms, including ARO, PSO,
GA, and GWO, to optimize the hyperparameters of a feed-forward neural network.
The experiments are conducted on the MNIST dataset, commonly used for image
analysis. Comparing the accuracy of models trained with these algorithms, with
and without dropout, reveals that bio-inspired algorithms improve deep learning
model accuracy. Incorporating bio-inspired algorithms in hyperparameter tuning
shows promise for optimizing deep learning models. By drawing inspiration from
nature, these algorithms enhance performance and generalization, demonstrating
their efficacy in deep learning hyperparameter tuning
Côte titre : MAI/0761 En ligne : https://drive.google.com/file/d/1WcfNow4t2G8wXoZTx2oNMTp_XM-CbHG0/view?usp=drive [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0761 MAI/0761 Mémoire Bibliothéque des sciences Anglais Disponible
Disponible
Titre : Elephant herding optimization for image begmentation Type de document : texte imprimé Auteurs : Harrouche,Oussama, Auteur ; Semcheddine,Moussa, Directeur de thèse Editeur : Setif:UFA Année de publication : 2020 Importance : 1 vol (70 f .) Format : 29 cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique Index. décimale : 004 - Informatique Côte titre : MAI/0421 En ligne : https://drive.google.com/file/d/1zjYqn6jyDSrFx_IRf7fBSJxUE5f7ZrGp/view?usp=shari [...] Format de la ressource électronique : Elephant herding optimization for image begmentation [texte imprimé] / Harrouche,Oussama, Auteur ; Semcheddine,Moussa, Directeur de thèse . - [S.l.] : Setif:UFA, 2020 . - 1 vol (70 f .) ; 29 cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique Index. décimale : 004 - Informatique Côte titre : MAI/0421 En ligne : https://drive.google.com/file/d/1zjYqn6jyDSrFx_IRf7fBSJxUE5f7ZrGp/view?usp=shari [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0421 MAI/0421 Mémoire Bibliothéque des sciences Français Disponible
DisponibleFast fuzzy c-menas for mr brain image segmentation / Serti,Chouaib
Titre : Fast fuzzy c-menas for mr brain image segmentation Type de document : texte imprimé Auteurs : Serti,Chouaib, Auteur ; Semcheddine,Moussa, Directeur de thèse Editeur : Setif:UFA Année de publication : 2019 Importance : 1 vol (53 f .) Format : 29 cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique Index. décimale : 004 Informatique Côte titre : MAI/0294 Fast fuzzy c-menas for mr brain image segmentation [texte imprimé] / Serti,Chouaib, Auteur ; Semcheddine,Moussa, Directeur de thèse . - [S.l.] : Setif:UFA, 2019 . - 1 vol (53 f .) ; 29 cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique Index. décimale : 004 Informatique Côte titre : MAI/0294 Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0294 MAI/0294 Mémoire Bibliothéque des sciences Français Disponible
DisponibleA Modified Black Widow Optimization Algorithm for Multilevel Thresholding Image Segmentation / Hocine Seif Eddine Lakhal
Titre : A Modified Black Widow Optimization Algorithm for Multilevel Thresholding Image Segmentation Type de document : texte imprimé Auteurs : Hocine Seif Eddine Lakhal, Auteur ; Khaled Ounis, Auteur ; Semcheddine,Moussa, Directeur de thèse Année de publication : 2022 Importance : 1 vol (65 f .) Format : 29cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Optimization
Bio-inspired algorithmsIndex. décimale : 004 - Informatique Résumé :
One of the most crucial aspects of image segmentation is multilevel thresholding. It works by partitioning
a digital image into multiple regions or sets. There are many methods for image segmentation but Multilevel
thresholding is the methods that easily performs this task. The key objective of any multilevel thresholding
method is to obtain the optimal thresholds on a target image for the region of interest. He problem then is to
find the best thresholds that properly segment each image. In this paper, we suggest a multilevel thresholding
algorithm using Modified Black Widow Optimization (MBWO) to find the best threshold configuration using
Otsu and Kapur as objective function. As for how this algorithm functions, it uses an opposite based
initialization to increase the quality of the initial population and the likelihood of starting with suitable
individuals, a delay in sexual cannibalism to prevent the loss of fit fathers, and adaptive crossover and
mutation probabilities to maintain a balance between exploration and exploitation.Côte titre : MAI/0696 En ligne : https://drive.google.com/file/d/1K9UQEO_lM6zbhXo7fhXEydPsI-J1EiBr/view?usp=share [...] Format de la ressource électronique : A Modified Black Widow Optimization Algorithm for Multilevel Thresholding Image Segmentation [texte imprimé] / Hocine Seif Eddine Lakhal, Auteur ; Khaled Ounis, Auteur ; Semcheddine,Moussa, Directeur de thèse . - 2022 . - 1 vol (65 f .) ; 29cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Optimization
Bio-inspired algorithmsIndex. décimale : 004 - Informatique Résumé :
One of the most crucial aspects of image segmentation is multilevel thresholding. It works by partitioning
a digital image into multiple regions or sets. There are many methods for image segmentation but Multilevel
thresholding is the methods that easily performs this task. The key objective of any multilevel thresholding
method is to obtain the optimal thresholds on a target image for the region of interest. He problem then is to
find the best thresholds that properly segment each image. In this paper, we suggest a multilevel thresholding
algorithm using Modified Black Widow Optimization (MBWO) to find the best threshold configuration using
Otsu and Kapur as objective function. As for how this algorithm functions, it uses an opposite based
initialization to increase the quality of the initial population and the likelihood of starting with suitable
individuals, a delay in sexual cannibalism to prevent the loss of fit fathers, and adaptive crossover and
mutation probabilities to maintain a balance between exploration and exploitation.Côte titre : MAI/0696 En ligne : https://drive.google.com/file/d/1K9UQEO_lM6zbhXo7fhXEydPsI-J1EiBr/view?usp=share [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0696 MAI/0696 Mémoire Bibliothéque des sciences Anglais Disponible
Disponible
Titre : Tuning deep learning parameters using bio-inspired algorithm Type de document : texte imprimé Auteurs : Aya Lakhdari, Auteur ; leila Djidel, Auteur ; Semcheddine,Moussa, Directeur de thèse Année de publication : 2022 Importance : 1 vol (67 f .) Format : 29cm Langues : Français (fre) Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique
InternetIndex. décimale : 004 Informatique Résumé :
There are several domains in which deep learning techniques are successfully applied.
The great success of deep learning depends on the great performance of models. Learning
rate, choice of activation function, network architectures..., all of these are deep learning
hyper-parameters that had an important role in improving the performance of models,
but there are no rules for choosing the right value of each hyper-parameter. In this
thesis, we will use some bio-inspired algorithms like genetic algorithm, particle swarm
optimization algorithm, black widow optimization algorithm, and grey wolf optimization
algorithm to optimize the CNN and LSTM models hyper-parameters. We have used two
data-set, the first is the MNIST data-set which concerned images and the second is the
20 newsgroups text data-set. And we have illustrated the results of the tests affected by
the bio-inspired algorithms and both random and grid searches. The results show that
bio-inspired algorithms give better accuracy results so it can be the way to tune deep
learning hyper-parameters.Côte titre : MAI/0675 En ligne : https://drive.google.com/file/d/1Jkj-sFnSsaxyLtTPAkh1Gz2q1HRUgbHK/view?usp=share [...] Format de la ressource électronique : Tuning deep learning parameters using bio-inspired algorithm [texte imprimé] / Aya Lakhdari, Auteur ; leila Djidel, Auteur ; Semcheddine,Moussa, Directeur de thèse . - 2022 . - 1 vol (67 f .) ; 29cm.
Langues : Français (fre)
Catégories : Thèses & Mémoires:Informatique Mots-clés : Informatique
InternetIndex. décimale : 004 Informatique Résumé :
There are several domains in which deep learning techniques are successfully applied.
The great success of deep learning depends on the great performance of models. Learning
rate, choice of activation function, network architectures..., all of these are deep learning
hyper-parameters that had an important role in improving the performance of models,
but there are no rules for choosing the right value of each hyper-parameter. In this
thesis, we will use some bio-inspired algorithms like genetic algorithm, particle swarm
optimization algorithm, black widow optimization algorithm, and grey wolf optimization
algorithm to optimize the CNN and LSTM models hyper-parameters. We have used two
data-set, the first is the MNIST data-set which concerned images and the second is the
20 newsgroups text data-set. And we have illustrated the results of the tests affected by
the bio-inspired algorithms and both random and grid searches. The results show that
bio-inspired algorithms give better accuracy results so it can be the way to tune deep
learning hyper-parameters.Côte titre : MAI/0675 En ligne : https://drive.google.com/file/d/1Jkj-sFnSsaxyLtTPAkh1Gz2q1HRUgbHK/view?usp=share [...] Format de la ressource électronique : Exemplaires (1)
Code-barres Cote Support Localisation Section Disponibilité MAI/0675 MAI/0675 Mémoire Bibliothéque des sciences Anglais Disponible
Disponible