Dear all,
Our next AI seminar is scheduled to be on May 31st (Friday).
There are 2 seminars on Friday this week: [Please note that they are in different locations]
Talk 1 :
Title: Training Machines to Know What They Don't Know
Time : 10-11 AM
Seminar Location: KEC 1003
Zoom link: https://oregonstate.zoom.us/j/91611213801?pwd=Wm9JSkN1eW84RUpiS2JEd0E5TEVkd…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…>
Talk 2 :
Title: Inverse constraint learning and risk averse reinforcement learning for safe AI
Time : 2-3 PM
Seminar Location: BEXL 320
Zoom link: https://oregonstate.zoom.us/j/91611213801?pwd=Wm9JSkN1eW84RUpiS2JEd0E5TEVkd…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…>
Speaker Details:
Pascal Poupart
Professor
David R. Cheriton School of Computer Science
University of Waterloo
Talk 1 Abstract: Training Machines to Know What They Don't Know
Modern machine learning predictors often suffer from overconfidence. I will explain why neural predictors with a softmax output layer exhibit arbitrarily high confidence away from the training data. I will describe a simple modification to the softmax layer to prevent this type of overconfidence. I will also describe how product-of-expert and mixture-of-expert approximations in Bayesian inference can lead to over or under confidence. Finally, I will describe a simple interpolation technique to enhance the calibration of predictions in distributed machine learning applications including federated learning.
Talk 2 Abstract: Inverse constraint learning and risk averse reinforcement learning for safe AI
In many applications of reinforcement learning (RL) and control, policies need to satisfy constraints to ensure feasibility, safety or thresholds about key performance indicators. However, some constraints may be difficult to specify. For instance, in autonomous driving, it is relatively easy to specify a reward function to reach a destination, but implicit constraints followed by expert human drivers to ensure a safe, smooth and comfortable ride are much more difficult to specify. I will present some techniques to learn soft constraints from expert trajectories in autonomous driving and robotics. I will also present an alternative to variance based on Gini deviation for risk-averse reinforcement learning.
Speaker Bio:
Pascal Poupart is a Professor in the David R. Cheriton School of Computer Science at the University of Waterloo (Canada). He is also a Canada CIFAR AI Chair at the Vector Institute and a member of the Waterloo AI Institute. He serves on the advisory board of the NSF AI Institute for Advances in Optimization (2022-present) at Georgia Tech. He served as Research Director and Principal Research Scientist at the Waterloo Borealis AI Research Lab at the Royal Bank of Canada (2018-2020). He also served as scientific advisor for ProNavigator (2017-2019), ElementAI (2017-2018) and DialPad (2017-2018). His research focuses on the development of algorithms for Machine Learning with application to Natural Language Processing and Material Discovery. He is most well-known for his contributions to the development of Reinforcement Learning algorithms. Notable projects that his research team are currently working on include inverse constraint learning, mean field RL, RL foundation models, Bayesian federated learning, uncertainty quantification, probabilistic deep learning, conversational agents, transcription error correction, sport analytics, adaptive satisfiability and material discovery for CO2 recycling.
Please watch this space for future AI Seminars :
https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…>
Note to AI Seminar course students: Credit will be given for both seminars
Rajesh Mangannavar,
Graduate Student
Oregon State University
----
AI Seminar Important Reminders:
-> For graduate students in the AI program, attendance is strongly encouraged
Dear all,
Our next AI seminar is scheduled to be on May 24th (Friday), 3-4 PM.
Please note the change in time for this week – It is from 3-4 PM this week.
Seminar Location: BEXL 320
Zoom link: https://oregonstate.zoom.us/j/91611213801?pwd=Wm9JSkN1eW84RUpiS2JEd0E5TEVkd…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…>
Deep Learning for Radio-Frequency Device Fingerprinting
Weng-Keen Wong
Professor
Computer Science
Oregon State University
Abstract:
Radio frequency (RF) device fingerprinting plays an important role in network security, enabling physical layer-based network access authentication and network device classification through the identification of devices from their transmitted RF signals. The use of deep learning for RF device fingerprinting has become prevalent in recent years as it has enabled the automated extraction of device-specific features and signatures solely from raw RF signals, without the need for expert domain knowledge to engineer informative features. This talk will describe two active areas of research related to RF device fingerprinting: 1) open-set detection for detecting unauthorized devices and 2) domain adaptation to deal with changes in channel conditions and environmental settings.
Speaker Bio:
Weng-Keen Wong received his Ph.D. and M. S. degrees in computer science from Carnegie Mellon University in 2004 and 2001 respectively. He received his B.Sc. degree from the University of British Columbia in 1997. He is currently a Professor in the School of Electrical Engineering and Computer Science at Oregon State University. From 2016-2018, he served as a program director at the National Science Foundation under the Robust Intelligence Program in the Division of Information and Intelligent Systems. His research areas are in data mining and machine learning, with specific interests in anomaly detection, deep learning, probabilistic graphical models, computational sustainability and human-in-the-loop learning.
Please watch this space for future AI Seminars :
https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…>
Rajesh Mangannavar,
Graduate Student
Oregon State University
----
AI Seminar Important Reminders:
-> For graduate students in the AI program, attendance is strongly encouraged
Dear all,
Our next AI seminar is scheduled to be on May 17th (Friday), 2-3 PM.
Seminar Location: BEXL 320
Student meeting: 3-3:30 pm, KEC 2057 (sign up here<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdocs.goog…>)
Zoom link: https://oregonstate.zoom.us/j/91611213801?pwd=Wm9JSkN1eW84RUpiS2JEd0E5TEVkd…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…>
Towards Automated Data Mining: Reinforcement Intelligence for Self-Optimizing Feature Engineering
Kunpeng Liu
Assistant Professor
Department of Computer Science
Portland State University
Abstract:
In recent years, data mining has achieved great success in enormous scenarios. As the foundation of data mining, feature engineering plays an essential role in comprehending and perceiving data. Successful feature engineering can remove irrelevant features, generate informative features, improve model performance, enhance generalization, and provide better interpretation and explanation. However, not all researchers and practitioners are experts in feature engineering, making the automation of feature engineering an indispensable demand. In this talk, I will first introduce what feature engineering is and why it is difficult to automate the feature engineering process. Then, I will focus on (1) automated feature selection (2) automated feature generation, and discuss how the framework of reinforcement learning can be adapted to solve these problems correspondingly. Finally, I will conclude the talk and present the big picture of developing intelligent, interpretable, and interactive automated data science systems.
Speaker Bio:
Kunpeng Liu is an assistant professor from the Department of Computer Science, Portland State University. His research interests are data mining and machine learning, especially in automated data science systems, with application to big data problems, including smart city, privacy-preserving machine learning, reasoning on recommender systems, and user behavior analysis. He is also interested in efficient RLHF and LLM inference.
Please watch this space for future AI Seminars :
https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…>
Rajesh Mangannavar,
Graduate Student
Oregon State University
----
AI Seminar Important Reminders:
-> For graduate students in the AI program, attendance is strongly encouraged
Dear all,
Our next AI seminar is scheduled to be on May 10th (Friday), 2-3 PM.
Seminar Location: BEXL 320
Student meeting: 3-3:30 pm, KEC 2057 (sign up here<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdocs.goog…>)
Zoom link: https://oregonstate.zoom.us/j/91611213801?pwd=Wm9JSkN1eW84RUpiS2JEd0E5TEVkd…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…>
When Sparse Data is All You Have: Learning World Models for Generalizable Planning and RL with a View Towards AI Assessment
Siddharth Srivastava
Associate Professor of Computer Science
School of Computing and Augmented Intelligence
Arizona State University
Abstract:
Can we build autonomous agents that learn generalizable knowledge and reliably solve previously unseen problems? In this talk I will present some of my group's recent research on neuro-symbolic learning for sequential decision-making problems that feature long-horizons, sparse rewards and vast differences between training and testing problems. I will discuss methods for learning and using abstractions to learn world models that can be easily transferred to new problem instances, often overshadowing the complexity of training instances. We will see how these methods can be used to invent symbolic vocabularies and learn logic-based world-models for robot task and motion planning without any human annotation or hand-written logic; how learning simple abstractions during Q-learning can vastly improve the performance of RL agents; and finally, how abstractions can help address the emerging problem of AI assessment.
Speaker Bio:
Siddharth Srivastava is an Associate Professor of Computer Science in the School of Computing and Augmented Intelligence at Arizona State University. He received his Ph.D. in Computer Science at the University of Massachusetts, Amherst, and did his postdoctoral research at UC Berkeley. His research focuses on safe and reliable taskable AI systems and AI assessment. He is a recipient of the NSF CAREER award, a Best Paper award at the International Conference on Automated Planning and Scheduling (ICAPS), an Outstanding Dissertation award at UMass Amherst, and a Best Final Year Thesis Award at IIT Kanpur. He served as conference Co-Chair for ICAPS 2019 and currently serves as an Associate Editor for the Journal of AI Research.
Please watch this space for future AI Seminars :
https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…>
Rajesh Mangannavar,
Graduate Student
Oregon State University
----
AI Seminar Important Reminders:
-> For graduate students in the AI program, attendance is strongly encouraged
Dear all,
Our next AI seminar is scheduled to be on May 3rd (Friday), 2-3 PM.
Seminar Location: BEXL 320
Student meeting: 3-3:30 pm, KEC 2057 (sign up here<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdocs.goog…>)
Zoom link: https://oregonstate.zoom.us/j/91611213801?pwd=Wm9JSkN1eW84RUpiS2JEd0E5TEVkd…<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Foregonsta…>
Teams of mobile robots orienteering in hazardous environments
Cory Simon
Associate Professor of Chemical Engineering
Oregon State University
Abstract:
Teams of mobile [aerial, ground, or aquatic] robots have applications in delivering resources, patrolling, information-gathering, pesticide application to crops, forest fire fighting, chemical plume source localization and mapping, and search-and-rescue. Some environments contain hazards e.g.,\ rough terrain or seas, strong winds, or adversaries capable of attacking robots. Then, the robots must coordinate their trails in the hazardous environment to (i) cooperatively achieve the team-level objective with robustness to robot failures and (ii) strike some balance between expected robot failures versus utility gained.
Herein, we pose the bi-objective, risky team orienteering problem, where (i) a team of robots are mobile within an environment, abstracted as a graph (nodes: locations, edges: spatial connections between locations); (ii) each node offers a reward to the team depending on the number of robots that visit it; (iii) the traversal of each edge in the graph imposes a risk of robot failure/destruction; and (iv) the two [often, conflicting] team objectives are to maximize the expected (a) team reward and (b) number of robots that survive the mission. We then use ant colony optimization to search for the Pareto-optimal set of robot team trail plans. A human decision-maker can then select the robot trail plans that balance, according to their values, the two [conflicting] objectives.
Speaker Bio:
Cory Simon learned the ropes of scientific research at the University of Akron, Virginia Tech, Okinawa Institute of Science and Technology, University of British Columbia, Lawrence Berkeley National Laboratory, École Polytechnique Fédérale de Lausanne, and Altius Institute for Biomedical Sciences. Simon interned in industry at Bridgestone Research (chemical engineering) and Stitch Fix (data science). He was a summer faculty research fellow at the Naval Information Warfare Center.
Please watch this space for future AI Seminars :
https://engineering.oregonstate.edu/EECS/research/AI<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fengineeri…>
Rajesh Mangannavar,
Graduate Student
Oregon State University
----
AI Seminar Important Reminders:
-> For graduate students in the AI program, attendance is strongly encouraged
