Abstract: My Robot can learn – using Reinforcement Learning to teach my Robot
A new star is rising at the machine learning horizon: reinforcement learning (RL). The concept entails an agent and an incentive-based training system. The agent learns via incentives and improves its behaviour – a self-learning system using simple rules – leading to artificial intelligence (AI). The talk covers an introduction to reinforcement learning and its combination with deep learning to achieve an AI system – a smart, intelligent bot! Annotating data to create a base model and its refinement through RL mechanisms brings us to the next level. Let’s build our self-learning robot!
Marcel is working as a Program Manager at Microsoft AI & Research in Germany. In the past his work was focused on data and high-scale systems. Nowadays, his focus is towards natural user interfaces and speech recognition. Thus, this brings a nice combination of a smart way of handling data and on the other side understanding interaction with humans and robots in a smart way. Besides this he enjoys giving talks, writing papers, do some coding and electronic experiments.
Olivia Klose (@oliviaklose) is a Software Development Engineer in the Commercial Software Engineering group at Microsoft. She is focussing on all analytics services on Microsoft Azure, in particular Hadoop (HDInsight), Spark and Machine Learning, and is a frequent speaker at German and international conferences, such as TechEd Europe, PASS Summit and Technical Summit. Prior to joining Microsoft, she studied Computer Science with Mathematics at the University of Cambridge, the Technical University of Munich and IIT Bombay. Here, she focussed on Machine Learning in Medical Imaging.