Difference between revisions of "First Level Theses"

From AIRWiki
Jump to: navigation, search
(New page: <!--==== Agents, Multiagent Systems, Agencies ====--> <!--==== BioSignal Analysis ====--> <!--==== Computer Vision and Image Analysis ====--> <!--==== E-Science ====--> ==== Machine Learni...)
 
Line 3: Line 3:
 
<!--==== Computer Vision and Image Analysis ====-->
 
<!--==== Computer Vision and Image Analysis ====-->
 
<!--==== E-Science ====-->
 
<!--==== E-Science ====-->
==== Machine Learning ====
+
<!--==== Machine Learning ====-->
 
+
<!--==== Ontologies and Semantic Web ====-->
 +
<!--==== Philosophy of Artificial Intelligence ====-->
 +
==== Robotics ====
 
{{Project template
 
{{Project template
|title=Reinforcement Learning in Poker
+
|title=Simulation of 6-DOF Robot Manipulator
|tutor=Marcello Restelli
+
|tutor=Marcello Restelli (restelli-AT-elet-DOT-polimi-DOT-it)
|description=In this years, Artificial Intelligence research has shifted its attention from fully observable environments such as Chess to more challenging partially observable ones such as Poker.
+
|description=The goal of this project is to develop a simulator for a 6-DOF robot manipulator, using the [http://www.ode.org/ ode] (open dynamics engine) library for simulating the rigid body dynamics. The project involves three different phases:
 +
* Building the physical model of the manipulator
 +
* Implementing the forward and inverse kinematic routines
 +
* Implementing the trajectory planning routines
 +
* Implementing the control modules
 +
* Implementing an interface to control the robot movements
  
Up to this moment research in this kind of environments, which can be formalized as Partially Observable Stochastic Games, has been more from a game theoretic point of view, thus focusing on the pursue of optimality and equilibrium, with no attention to payoff maximization, which may be more interesting in many real-world contexts.
+
This project allows to put into practice what has been explained during the first part of the course of Robotics.
  
On the other hand Reinforcement Learning techniques demonstrated to be successful in solving both fully observable problems, single and multi-agent, and single-agent partially observable ones, while lacking application to the partially observable multi-agent framework.
+
The project can be turned into a thesis, by using the simulated manipulator to perform some learning experiments.
 
+
This research aims at studying the solution of Partially Observable Stochastic Games, analyzing the possibility to combine the Opponent Modeling concept with the well proven Reinforcement Learning solution techniques to solve problems in this framework, adopting Poker as testbed.
+
 
|start=Anytime
 
|start=Anytime
|number=2
+
|number=2-3
|cfu=5
+
|cfu=10-15
|image=PokerPRLT.png}}
+
|image=puma6dof1.jpg}}
 
+
<!--==== Ontologies and Semantic Web ====-->
+
<!--==== Philosophy of Artificial Intelligence ====-->
+
<!--==== Robotics ====-->
+
 
<!--==== Soft Computing ====-->
 
<!--==== Soft Computing ====-->

Revision as of 15:06, 29 September 2008

Robotics


Title: Simulation of 6-DOF Robot Manipulator
Puma6dof1.jpg
Description: The goal of this project is to develop a simulator for a 6-DOF robot manipulator, using the ode (open dynamics engine) library for simulating the rigid body dynamics. The project involves three different phases:
  • Building the physical model of the manipulator
  • Implementing the forward and inverse kinematic routines
  • Implementing the trajectory planning routines
  • Implementing the control modules
  • Implementing an interface to control the robot movements

This project allows to put into practice what has been explained during the first part of the course of Robotics.

The project can be turned into a thesis, by using the simulated manipulator to perform some learning experiments.

Tutor: Marcello Restelli (restelli-AT-elet-DOT-polimi-DOT-it)
Start: Anytime
Number of students: 2-3
CFU: 10-15