Français Anglais
Accueil Annuaire Plan du site
Home > Research results > Dissertations & habilitations
Research results
Ph.D de

Ph.D
Group : Learning and Optimization

Recurrent Neural Networks and Reinforcement Learning

Starts on 01/10/2016
Advisor : OLLIVIER, Yann

Funding : Contrat doctoral spécifique normalien ou polytechnicien
Affiliation : Université Paris-Sud
Laboratory : LRI - AO

Defended on 07/10/2019, committee :
Directeur de thèse :
- OLLIVIER Yann, CNRS

Rapporteurs :
- M. Joan BRUNA, Université de New York
- M. Pascal VINCENT, Université de Montréal

Examinateurs :
- Mme Anne VILNAT, Université Paris-Sud
- M. Francis BACH, École Normale Supérieure
- M. Jean-Philippe VERT, Mines ParisTech

Research activities :

Abstract :
An intelligent agent immerged in its environment must be able to both
understand and interact with the world. Understanding the environment requires
processing sequences of sensorial inputs. Interacting with the environment
typically involves issuing actions, and adapting those actions to strive
towards a given goal, or to maximize a notion of reward. This view of a two
parts agent-environment interaction motivates the two parts of this thesis: recurrent
neural networks are powerful tools to make sense of complex and diverse
sequences of inputs, such as those resulting from an agent-environment
interaction; reinforcement learning is the field of choice to direct the
behavior of an agent towards a goal. This thesis aim is to provide theoretical
and practical insights in those two domains. In the field of recurrent
networks, this thesis contribution is twofold: we introduce two new,
theoretically grounded and scalable learning algorithms that can be used online.
Besides, we advance understanding of gated recurrent networks, by examining their
invariance properties. In the field of reinforcement learning, our main
contribution is to provide guidelines to design time discretization robust
algorithms. All these contributions are theoretically grounded, and backed up
by experimental results.

Ph.D. dissertations & Faculty habilitations
DESIGNING INTERACTIVE TOOLS FOR CREATORS AND CREATIVE WORK
Creative work has been at the core of research in Human-Computer Interaction (HCI). I describe the results of a series of studies that look at how creators work, where creators include artists with years of professional practice, as well as learners, or novices and casual makers. My research focuses on three creation activities: drawing, physical modeling, and music composition. For these activities, I examine how artists switch between representations and how these representations evolve throughout their creative process, from early sketches to fine-grained forms or structured vocabularies. I present interactive systems that enrich their workflow (i) by extending their computer tools with physical user interfaces, or (ii) by making physical materials interactive. I also argue that sketch-based representations can allow for user interfaces that are more personal and less rigid. My presentation will reflect on lessons and limitations of this work and discuss challenges for future design-support tools.

INCREASING THE BANDWIDTH OF INTERACTIVE VISUALIZATIONS, USING COMPLEX DISPLAY ENVIRONMENTS AND TARGETED DESIGNS
Interactive visualizations combine human computer interaction, visual design, perception theory, as well as data processing methods in order to propose visual data representations that amplify cognition, and aid data exploration and understanding. We can consider visualization as a communication medium or channel between humans and their data. The higher the communication bandwidth (the data that can be communicated and understood), the more effective the visualization is. My research attempts to increase the bandwidth of this communication channel in the following two ways. (i) First, by moving away from traditional desktops towards larger displays that can both render larger amounts of data and can accommodate multiple viewers. (ii) And second, by designing and studying appropriate visual representations that show salient information. In my presentation I will describe my work on these topics, the challenges it tries to address, and discuss the methodology and inspiration behind this research.

MODéLISATION DE SYSTèME PHYSIQUES PAR APPRENTISSAGE STATISTIQUE PROFOND