m

F Nous contacter

0

Documents  68T05 | enregistrements trouvés : 70

O

-A +A

P Q

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research talks;Computer Science

In this presentation based on on-line demonstrations of algorithms and on the examination of several practical examples, I will reflect on the problem of modeling a detection task in images. I will place myself in the (very frequent) case where the detection task can not be formulated in a Bayesian framework or, rather equivalently that can not be solved by simultaneous learning of the model of the object and that of the background. (In the case where there are plenty of examples of the background and of the object to be detected, the neural networks provide a practical answer, but without explanatory power). Nevertheless for the detection without "learning", I will show that we can not avoid building a background model, or possibly learn it. But this will not require many examples.

Joint works with Axel Davy, Tristan Dagobert, Agnes Desolneux, Thibaud Ehret.
In this presentation based on on-line demonstrations of algorithms and on the examination of several practical examples, I will reflect on the problem of modeling a detection task in images. I will place myself in the (very frequent) case where the detection task can not be formulated in a Bayesian framework or, rather equivalently that can not be solved by simultaneous learning of the model of the object and that of the background. (In the case ...

65D18 ; 68U10 ; 68T05

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research talks;Computer Science

68Wxx ; 68P05 ; 68M11 ; 68U20 ; 68Q80 ; 68T05 ; 94A60 ; 94A08

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research talks;Combinatorics;Computer Science;Probability and Statistics

A non-backtracking walk on a graph is a directed path such that no edge is the inverse of its preceding edge. The non-backtracking matrix of a graph is indexed by its directed edges and can be used to count non-backtracking walks of a given length. It has been used recently in the context of community detection and has appeared previously in connection with the Ihara zeta function and in some generalizations of Ramanujan graphs. In this work, we study the largest eigenvalues of the non-backtracking matrix of the Erdos-Renyi random graph and of the Stochastic Block Model in the regime where the number of edges is proportional to the number of vertices. Our results confirm the "spectral redemption" conjecture that community detection can be made on the basis of the leading eigenvectors above the feasibility threshold. A non-backtracking walk on a graph is a directed path such that no edge is the inverse of its preceding edge. The non-backtracking matrix of a graph is indexed by its directed edges and can be used to count non-backtracking walks of a given length. It has been used recently in the context of community detection and has appeared previously in connection with the Ihara zeta function and in some generalizations of Ramanujan graphs. In this work, we ...

05C50 ; 05C80 ; 68T05 ; 91D30

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- xiii; 538 p.
ISBN 978-0-941743-64-8

Localisation : Colloque 1er étage (ANTI)

analyse multivariée # analyse factorielle # acquisition des connaissances # système expert # apprentissage automatique # ordinateur neuronal

62-06 ; 62-07 ; 68Q32 ; 68Txx ; 68T05 ; 68T30

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- xi; 343 p.
ISBN 978-3-642-24872-6

Lecture notes in artificial intelligence , 6992

Localisation : Colloque 1er étage (PISC)

théorie de la décision # décision multicritère # informatique # programmation

90-06 ; 68-06 ; 91-06 ; 68T05 ; 90B50 ; 90C29 ; 91B06 ; 00B25

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- xiv; 370 p.
ISBN 978-0-8218-4380-2

DIMACS series in discrete mathematics and theoretical computer science , 0069

Localisation : Collection 1er étage

combinatoires # informatique # chimie # graphe

68R10 ; 68T05 ; 68T35 ; 05C35 ; 05C30 ; 05C62 ; 05-06 ; 68-06 ; 00B25

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- 132 p.

C.I.M. , 0021

Localisation : Colloque 1er étage (COIM)

système complexe # logique floue # réseau neuronal # calcul évolutif # complexité

68T05 ; 93C42

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- 224 p.
ISBN 978-0-8218-3483-1

Contemporary mathematics , 0349

Localisation : Collection 1er étage

théorie des groupes # groupe de permutation # groupe non-abélien # théorie quantique # informatique # algorithme # groupe d'automorphisme # calcul quantique

20B40 ; 20E05 ; 20F28 ; 81P68 ; 68Q05 ; 68Q17 ; 68Q42 ; 68Q45 ; 68T05

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.


ISBN 978-0-8186-0742-4

Localisation : Colloque 1er étage (PARI)

algorithme # application biomédicale # architecture # automate # bord # classification des données # connaissance de base de reconnaissance de forme # inférence et apprentissage # limite # tomographie # traitement des images # traitement du signal # vision

68Q20 ; 68Q25 ; 68T05 ; 68T10 ; 68T30 ; 68Txx ; 68U05 ; 68U10 ; 68U30 ; 68Uxx

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.


ISBN 978-0-521-65263-6

Publications of the newton institute

Localisation : Colloque 1er étage (CAMB)

analyse bayesienne # analyse de prototype en ligne # analyse en composante principale # analyse statistique # apprentissage # apprentissage en ligne # approximation stochastique # enseignement en ligne # réseau de neurone

68T05 ; 82-06

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- Pp. 453-909
ISBN 978-0-904933-61-1

Localisation : Colloque 1er étage (AVIG)

application # contrôle # gestion # industrie des systèmes experts # intelligence artificielle # outil et technique de construction des systèmes experts # surveillance # système expert # système informatique # sécurité # télécommunication # économie des systèmes experts # électrotechnique

68T05 ; 68T35 ; 68Txx ; 68U07 ; 68U20 ; 68U30 ; 68Uxx

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

- 449 p.
ISBN 978-0-904933-61-1

Localisation : Colloque 1er étage (AVIG)

application # contrôle # gestion # industrie des systèmes experts # intelligence artificielle # outil et technique de construction des systèmes experts # surveillance # système expert # système informatique # sécurité # télécommunication # économie des systèmes experts # électrotechnique

68T05 ; 68T35 ; 68Txx ; 68U07 ; 68U20 ; 68U30 ; 68Uxx

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.


ISBN 978-1-55860-300-4

Localisation : Colloque 1er étage (CHAM)

apprentissage machine # langage naturel # paneau # panel # physique naive # planning # programmation logique # qualitatif # raisonnement # reseau neuronnaux # robotique et vision # vidéo # vision

68T05 ; 68T25 ; 68T27 ; 68T30 ; 68T35 ; 68Txx ; 68Uxx

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.


ISBN 978-1-55860-300-4

Localisation : Colloque 1er étage (CHAM)

AI distribuée # automate # automatique # intelligence artificielle # modèle co-positif # problème de satisfaction des contraintes # représentation des connaissances # système intelligent de Tutoring # technologie des connaissances de base

68T05 ; 68T20 ; 68T30 ; 68T35 ; 68Txx

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.


ISBN 978-0-8186-0742-4

Localisation : Colloque 1er étage (PARI)

algorithme # application biomédicale # architecture # automate # bord # classification des données # connaissance de base de reconnaissance de forme # inférence et apprentissage # limite # tomagraphie # traitement des images # traitement du signal # vision

68Q20 ; 68Q25 ; 68T05 ; 68T10 ; 68T30 ; 68Txx ; 68U05 ; 68U10 ; 68U30 ; 68Uxx

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.


ISBN 978-0-521-47278-4

Publications of the Newton Institut , 0004

Localisation : Colloque 1er étage (CAMB)

contrôle de comportement guidé visuellement # guidage de robot visuel à partir de stéréo non-calibrée # traquage et contrôle de voitures jouets # traquage visuel # vision et exploration basées sur un modèle # vision informatique en temps réel

68T05 ; 68T10 ; 68Txx ; 68U10

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research talks

Analysis of algorithms in noisy high-dimensional probabilistic problems poses many current challenges. In a subclass of these problems the corresponding challenges can be overcome with the help of a method coming from statistical mechanics. I will review some of the related recent work together with progress on rigorous justification of the corresponding results.

68T05 ; 62P35 ; 68W25

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research talks;Computer Science;Geometry

Neural networks consist of a variegate class of computational models, used in machine learning for both supervised and unsupervised learning. Several topologies of networks have been proposed in the literature, since the preliminary work from the late 50s, including models based on undirected probabilistic graphical models, such as (Restricted) Boltzmann Machines, and on multi-layer feed-forward computational graphs. The training of a neural network is usually performed by the minimization of a cost function, such as the negative log-likelihood. During the talk we will review alternative geometries used to describe the space of the functions encoded by a neural network, parametrized by its connection weights, and the implications on the optimization of the cost function during training, from the perspective of Riemannian optimization. In the first part of the presentation, we will introduce a probabilistic interpretation for neural networks, which goes back to the work of Amari and coauthors from the 90s, and which is based on the use of the Fisher-Rao metric studied in Information Geometry. In this framework, the weights of a Boltzmann Machine, and similarly for feed-forward neural networks, are interpreted as the parameters of a (joint) statistical model for the observed, and possibly latent, variables. In the second part of the talk, we will review other approaches, motivated by invariant principles in neural networks and not explicitly based on probabilistic models, to the definition of alternative geometries for the space of the parameters of a neural network. The use of alternative non-Euclidean geometries has direct impact on the training algorithms, indeed by modeling the space of the functions associated to a neural network as a Riemannian manifold determines a dependence of the gradient on the choice of metric tensor. We conclude the presentation by reviewing some recently proposed training algorithm for neural networks, based on Riemannian optimization algorithms. Neural networks consist of a variegate class of computational models, used in machine learning for both supervised and unsupervised learning. Several topologies of networks have been proposed in the literature, since the preliminary work from the late 50s, including models based on undirected probabilistic graphical models, such as (Restricted) Boltzmann Machines, and on multi-layer feed-forward computational graphs. The training of a neural ...

53B21 ; 65K10 ; 68T05 ; 92B20

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research schools;Computer Science;Probability and Statistics

In this talk, I will introduce the classical theory of multi-armed bandits, a field at the junction of statistics, optimization, game theory and machine learning, discuss the possible applications, and highlights the new perspectives and open questions that they propose We consider competitive capacity investment for a duopoly of two distinct producers. The producers are exposed to stochastically fluctuating costs and interact through aggregate supply. Capacity expansion is irreversible and modeled in terms of timing strategies characterized through threshold rules. Because the impact of changing costs on the producers is asymmetric, we are led to a nonzero-sum timing game describing the transitions among the discrete investment stages. Working in a continuous-time diffusion framework, we characterize and analyze the resulting Nash equilibrium and game values. Our analysis quantifies the dynamic competition effects and yields insight into dynamic preemption and over-investment in a general asymmetric setting. A case-study considering the impact of fluctuating emission costs on power producers investing in nuclear and coal-fired plants is also presented. In this talk, I will introduce the classical theory of multi-armed bandits, a field at the junction of statistics, optimization, game theory and machine learning, discuss the possible applications, and highlights the new perspectives and open questions that they propose We consider competitive capacity investment for a duopoly of two distinct producers. The producers are exposed to stochastically fluctuating costs and interact through aggregate ...

62L05 ; 68T05 ; 91A26 ; 91A80 ; 91B26

... Lire [+]

Déposez votre fichier ici pour le déplacer vers cet enregistrement.

Research schools;Computer Science

In this talk I will present some recent developments in model-free reinforcement learning applied to large state spaces, with an emphasis on deep learning and its role in estimating action-value functions. The talk will cover a variety of model-free algorithms, including variations on Q-Learning, and some of the main techniques that make the approach practical. I will illustrate the usefulness of these methods with examples drawn from the Arcade Learning Environment, the popular set of Atari 2600 benchmark domains. In this talk I will present some recent developments in model-free reinforcement learning applied to large state spaces, with an emphasis on deep learning and its role in estimating action-value functions. The talk will cover a variety of model-free algorithms, including variations on Q-Learning, and some of the main techniques that make the approach practical. I will illustrate the usefulness of these methods with examples drawn from the Arcade ...

68Q32 ; 91A25 ; 68T05

... Lire [+]

Z