Machine Learning for Autonomous Robotic Agents
نویسندگان
چکیده
We present some results of our research in the field of Machine Learning applied to robotics problems. In particular we have investigated on: (i) the application of Learning Classifier Systems to the synthesis of robot controllers; (ii) learning of fuzzy controllers; (iii) learning of purposeful representations of the environment; (iv) and the application of versions of Q-learning to robot training. Experimental results suggest that Machine Learning techniques might soon have an important impact on realistic robotic applications.
منابع مشابه
Autonomous Interpretation of Elevator
The objective of the work presented in this paper is develop a method to interpret interior elevator panels for robot navigation. We outline a three-level hierarchical approach applies principles from both supervised and unsupervised machine learning techniques. In particular, we will show the development of an algorithm that accepts an elevator panel image paired with a desired floor location ...
متن کاملActive Learning Agents with Artificial Neural Brains
Two case studies on active learning agents are described. In the first case, the agent is a robot arm whose task is to grasp a rolling ball. It learns from self-generated examples by initiating conversations with the external teacher. In the second case, the agent is an autonomous mobile robot that works in cooperation with other mobile agents. The mobile robots learn group behavior by actively...
متن کاملAre Autonomous Mobile Robots Able to Take Over Construction? A Review
Although construction has been known as a highly complex application field for autonomous robotic systems, recent advances in this field offer great hope for using robotic capabilities to develop automated construction. Today, space research agencies seek to build infrastructures without human intervention, and construction companies look to robots with the potential to improve construction qua...
متن کاملEvaluating Pinch Quality of Underactuated Robotic Hands
Underactuated robotic hands are favorable for some applications due to its simplicity in mechanism and control compared with fully actuated hand. Incorporating sensors makes it possible for a hand to accomplish low level task autonomously. This project aims at analyzing pinch quality of an underactuated hand that was developed for underwater teleoperation with 8 sensor channels on each finger, ...
متن کاملCollaborative Q(λ) Reinforcement Learning Algorithm - a Promising Robot Learning Framework
This paper presents the design and implementation of a new reinforcement learning (RL) based algorithm. The proposed algorithm, ) ( CQ (collaborative ) ( Q ) allows several learning agents to acquire knowledge from each other. Acquiring knowledge learnt by an agent via collaboration with another agent enables acceleration of the entire learning system; therefore, learning can be utilized more...
متن کامل