Efficient sequential learning in structured and constrained environments
L'avantage principal des méthodes d'apprentissage non-paramétriques réside dans le fait que la nombre de degrés de libertés du modèle appris s'adapte automatiquement au nombre d'échantillons. Ces méthodes sont cependant limitées par le "fléau de la kernelisation": appre...
Main Author: | Calandriello, Daniele |
---|---|
Other Authors: | Lille 1 |
Language: | en |
Published: |
2017
|
Subjects: | |
Online Access: | http://www.theses.fr/2017LIL10216/document |
Similar Items
-
Multi-lingual dependency parsing : word representation and joint training for syntactic analysis
by: Dehouck, Mathieu
Published: (2019) -
Reinforcement learning for dialogue systems optimization with user adaptation
by: Carrara, Nicolas
Published: (2019) -
Flexible and efficient Gaussian process models for machine learning
by: Snelson, Edward Lloyd
Published: (2007) -
Efficient reinforcement learning in agents through embodiment-based modular decomposition
by: Jacob, David
Published: (2005) -
AgentP : a learning classifier system with associative perception in maze environments
by: Zatuchna, Zhanna V.
Published: (2005)