NS Seminar

Date and Location

May 20, 2019 - 11:00am to 12:00pm
Bldg 434, room 122

Abstract

Model Based Reinforcement Learning (presented by Alon Albalak, Computer Science)

Alon will take us through his current work in reinforcement learning and control theory. He will start with the ideas on state space models and deep learning as laid out in Rangapuram et al (2018) and then discuss how he is working on expanding the research.  

 

Advancing Deep Space Exploration via Simulators and Computer Vision (presented by James Bird, Computer Science)

James will present his current and future research on using machine learning techniques and computer vision to bolster the efforts of deep space exploration. He will discuss the current state of space exploration, as well as the problems inhibiting the field's advancement. He will then introduce novel technologies in development in the Physics Department at UCSB and the role he is playing in that development: The development of algorithms for celestial object recognition and ranking. Finally, he will propose his project for the use of NSF Competitive Innovation Incentive Funds, and the development of a simulator to train his models.