The last few years have seen multiple major successes of deep learning. From beating professionals at games like Go, to fast detection of cancer, classification of complex images, and generation of captions for images with incomplete information. In classification and reinforcement learning, deep learning indeed has outperformed all existing machine learning and model-based methods. Fueled by rapid advances in parallel computing and GPUs, new impressive results are achieved almost on a daily basis by deep learning methods in a vast array of applications from finance to self-driving cars. In the midst of this excitement however, there are fundamental questions that remain unanswered: For example, recent results have also shown that deep learning can be highly susceptible to noise and can misclassify images with very high confidence that are very easy for humans to process. If deep learning is going to make its way into self-driving cars and other autonomous systems, it is justifiable to ask just how stable and robust can deep learning based control and decision making algorithms be? Indeed, one could ask a even deeper question: What is the role of feedback in an increasingly data-driven world? This we believe is a question that the controls community can significantly contribute to answering. In the view of this vision, the objective of this workshop is to:
- Introduce the theory, algorithms, and implementation of Deep Learning to the audience, including conducting tutorials on using deep learning software (Google Tensor Flow). Embedded in this introduction, is a compact introduction to machine learning in general
- Have a candid conversation about the connection of deep learning with the Neural Network theory that has been developed in the controls community in the late 90s-early 2000s, and the larger controls efforts in system identification, adaptive control, and Kalman filtering
- Bring together experts from deep learning and machine learning with control theory experts
- Hold a discussions on what critical roles feedback control and estimation play in a world dominated by data-driven learning
The workshop is being organized by experts who have a track record of publications in control theory as well as highly competitive machine learning venues. Leveraging our contacts (and the location of the conference), the organizers will invite and bring together experts from Machine Learning to CDC if the workshop proposal is accepted. Many of these folks would typically not visit CDC, we hope that this won’t be their last CDC.
On the other hand, the workshop will also introduce deep learning to the CDC audience, and reveal to a large extent that while the fundamental concepts remain rooted in backpropagation as was studied in the controls community, the recent implementations differentiate and innovate in unique and powerful ways. To ensure that we leave our audience with a strong place to begin their journey in deep learning, we will execute two hour-long tutorials complete with sample code: 1. Classification with Google Tensor Flow package, 2. Deep Reinforcement Learning with Keras RL.
The target audience for this workshop are control theorists and practicing controls engineers. We will assume only familiarity with MATLAB and some familiarity with Python programming language (we will circulate material beforehand to the registrants to prime them in Python). Our presentations will be self-contained and leverage material that instructors have developed in teaching courses on related topics UIUC or their organizations. In addition, we will provide a repository of relevant links that help the audience navigate the wide variety of literature and resources on the internet on deep learning.
|8:00 - 8:20 AM
||Introduction and Welcome.
|8:20 - 8:40 AM
||Some results with Deep Learning
|8:40 - 9:00 AM
||What's the big deal, economically?
|9:00 - 11:00 AM
||Module 1: Deep learning algorithms and software.
Girish Chowdhary, Alex Schwing, Hassan Kingravi, Moitreya Chatterji
|11:00 - 12:00 PM
Deep Neural Networks:
Function Approximation and Classification.
|12:00 - 1:00 PM
|1:00 - 2:30 PM
||Module 2: Part 1:
Deep Reinforcement Learning.
|Girish Chowdhary, Alex Schwing, Anay Pattanaik, Moitreya Chatterji, Girish Joshi
|2:30 - 3:00 PM
|3:00 - 4:00 PM
Deep Recurrent Networks and Evolving Gaussian Processes
|Hassan Kingravi, Girish Chowdhary, Moitreya Chatterji, Joshua Whitman
|4:00 - 5:00 PM
||Discussion about open problems and challenges facing the community.
This workshop is proposed as a full day workshop. Experts from deep learning will be invited to provide insights and supplement the tutorials provided by the core organizing team. The final agenda of the workshop will follow the general outline described above, but may be organized as a series of short talks focusing on each of the key topics. The software tutorials will require python programming, and participants will be requested to bring their computers. Information on installing tensor flow and other required software packages will be provided in advance of the workshop.
Workshop Team: The core proposing team for this workshop are individuals with proven track records in publication in both machine learning and controls venues. We are a team that spans academia, industry, and startups working at the intersection of control theory, machine learning, and robotics. In addition, we will invite leading members of the deep learning community for invited presentations for this workshop if it is accepted. We have interest in participating in the workshop from researchers at major car companies and other companies interested in applications of deep learning. The biographies of the core proposing team follow:
Girish Chowdhary is an assistant professor and the director of the Distributed Autonomous Systems lab at the University of Illinois at Urbana-Champaign. He holds appointments with Agricultural and Biological engineering, Coordinated Science Lab, Aerospace Engineering, and the Beckman Institute. He holds a PhD (2010) from Georgia Institute of Technology. He was a postdoc at the Laboratory for Information and Decision Systems (LIDS) of the Massachusetts Institute of Technology, and an assistant professor at Oklahoma State University for three years prior to moving to UIUC. Girish was a member of the team that launched the first UAV project at the German Aerospace Center's (DLR) in 2003, where he worked for three years after completing his BS (2003) in Aerospace Engineering from the Royal Melbourne Institute of Technology in Melbourne, Australia. Girish's ongoing research interest is in theoretical insights and practical algorithms for adaptive autonomy. He is the author of over 90 peer reviewed publications in adaptive control, machine learning, and robotics. His latest paper at NIPS 2016 (a leading machine learning conference) laid the foundations for integrating notions from feedback estimation in predictive machine learning. He is the winner of the Air Force Young Investigator Award, AIAA GNC best paper award, and the Aerospace Guidance and Controls Systems Committee Dave Ward Memorial award for significant contributions to adaptive autopilots. He is an associate fellow of AIAA and senior member of IEEE.
Alex Schwing is an assistant professor in the ECE department of the University of Illinois at Urbana-Champaign (UIUC). Before joining UIUC he was a postdoctoral fellow at the University of Toronto after having received a PhD from ETH Zurich. His research interests are in the general areas of machine learning, deep learning, and computer vision. He is particularly interested in algorithms for prediction with and learning of nonlinear, multivariate and structured distributions, and their application in numerous tasks, e.g., for 3D scene understanding from a single image.
Hassan A. Kingravi is a Research Scientist with Pindrop, the leading provider of intelligent fraud detection systems in the telephony space. He earned his PhD in Electrical and Computer Engineering from the Georgia Institute of Technology, where his thesis made contributions to both the foundations of kernel-based machine learning and its applications to control theory. He joined Pindrop after a postdoc at Oklahoma State University. He is the author of nearly 40 peer-reviewed publications in computer vision, machine learning and control theory. In industry, he has developed systems for automated fraud discovery, identification of scams from robocalls, and holistic authentication for the financial sector.
Co-founder of EarthSense
Chinmay Soman is the co-founder of EarthSense, Inc. - a robotics and machine-learning startup developing autonomous low-cost robotic solutions for agriculture that complement existing high-cost agricultural equipment. Chinmay’s background is in agricultural sustainability (as a US National Science Foundation Postdoctoral Fellow), bioinformatics, and biotechnology. Based on his familiarity with real-world large-scale agricultural systems, Chinmay’s company is counting on Deep Learning to deliver unprecedented new solutions to these challenges, as well as a clear understanding of the complexity and difficulty of creating solutions that can deliver the necessary value in the harsh, uncertain, dynamic conditions in real-world environments.
Grad Student in Electrical and Computer Engineering
For more information please click here.
Grad Student in CSL
Anay Pattanaik is a second year graduate student working with Prof. Girish Chowdhary in Distributed Autonomous Systems Lab (DAS Lab). He graduated from Indian Institute of Technology Kanpur (IIT Kanpur) in 2016 with a B.Tech. degree in Aerospace Engineering. His current research interests lie in Reinforcement Learning and learning based control systems.
Anwesa gruduated with a Bachelor’s degree in Electronics and Communication Engineering from NIT Durgapur, India, in 2016. Prior to joining UIUC, Anwesa has worked as a researcher in the Advanced Robotics Centre, National University of Singapore for one year. Her areas of interest include computer vision, machine learning and robotics.
Aerospace Engineering (Autonomous Systems)
For more information please click here.
Undergrad in ABE Off-Road Equipment Engineering.
Computer Science minor.
For more information please click here.