Select a calendar:
Filter October Events by Event Type:
SUNMONTUEWEDTHUFRISAT
Conferences, Lectures, & Seminars
Events for October
-
CS Colloquium: Jonathan Kelly (University of Toronto) - Keeping Your Distances: A Distance-Geometric Perspective on Inverse Kinematics
Tue, Oct 04, 2022 @ 03:30 PM - 04:50 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Jonathan Kelly, University of Toronto
Talk Title: Keeping Your Distances: A Distance-Geometric Perspective on Inverse Kinematics
Series: Computer Science Colloquium
Abstract: In this talk, I will discuss recent work in my group on the problem of inverse kinematics (IK): finding joint angles that achieve a desired robot manipulator end-effector pose. A wide range of IK solvers exist, the majority of which operate on joint angles as parameters. Because the problem is highly nonlinear, these solvers are prone to local minima (among other troubles). I will introduce an alternate formulation of IK based on distance geometry, where a robot model is defined in terms of distances between rigidly-attached points. This alternative geometric description of the kinematics reveals an elegant equivalence between IK and the problem of low-rank Euclidean distance matrix completion. We use this connection to implement two novel solutions to IK for various articulated robots. The first is a Riemannian optimization-based approach which leverages the structure of the EDM manifold. The second solves a series of convex semidefinite relaxations of the distance-geometric problem. Both methods outperform many existing solvers on a variety of IK problems, some of which incorporate collision avoidance and joint limit constraints. Finally, I will describe a learned IK solver we have recently developed that is able to quickly generate sets of diverse approximate IK results for many different manipulators.
Prof. Kelly will give his talk in person at RTH 115 and we will also host the talk over Zoom.
Register in advance for this webinar at:
https://usc.zoom.us/webinar/register/WN_ShCRcm1iTrq6ubaaY-8P5Q
After registering, attendees will receive a confirmation email containing information about joining the webinar.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Prof. Jonathan Kelly leads the Space & Terrestrial Autonomous Robotic Systems (STARS) Laboratory at the University of Toronto Institute for Aerospace Studies. His group carries out research primarily in the areas of robotic perception, planning, and manipulation. Prof. Kelly holds a Canada Research Chair (Tier II) in Collaborative Robotics and was Dean's Catalyst Professor (an early-career award for research excellence) from 2018 to 2021. Prior to joining the University of Toronto, he was a postdoctoral fellow in CSAIL at MIT, working with Prof. Nick Roy. He received his PhD degree in 2011 from the University of Southern California under the supervision of Prof. Gaurav Sukhatme. At USC, he was a member of the first cohort of Annenberg Fellows. Although he lives in Toronto, he still sneaks back to Los Angeles to go scuba diving whenever he can.
Host: Stefanos Nikolaidis
Webcast: https://usc.zoom.us/webinar/register/WN_ShCRcm1iTrq6ubaaY-8P5QLocation: Ronald Tutor Hall of Engineering (RTH) - 115
WebCast Link: https://usc.zoom.us/webinar/register/WN_ShCRcm1iTrq6ubaaY-8P5Q
Audiences: Everyone Is Invited
Contact: Department of Computer Science
-
Machine Learning Center Seminar: Furong Huang (University of Maryland) - Trustworthy Machine Learning in Complex Environments
Wed, Oct 12, 2022 @ 10:00 AM - 11:30 AM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Furong Huang, University of Maryland
Talk Title: Trustworthy Machine Learning in Complex Environments
Series: Machine Learning Seminar Series hosted by USC Machine Learning Center
Abstract: With the burgeoning use of machine learning models in an assortment of applications, there is a need to rapidly and reliably deploy models in a variety of environments. These trustworthy machine learning models must satisfy certain criteria, namely the ability to: (i) adapt and generalize to previously unseen worlds although trained on data that only represent a subset of the world, (ii) allow for non-iid data, (iii) be resilient to (adversarial) perturbations, and (iv) conform to social norms and make ethical decisions.
In this talk, towards trustworthy and generally applicable intelligent systems, I will cover some reinforcement learning algorithms that achieve fast adaptation by guaranteed knowledge transfer, principled methods that measure the vulnerability and improve the robustness of reinforcement learning agents, and ethical models that make fair decisions under distribution shifts.
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Furong Huang is an Assistant Professor of the Department of Computer Science at University of Maryland. She works on statistical and trustworthy machine learning, reinforcement learning, graph neural networks, deep learning theory and federated learning with specialization in domain adaptation, algorithmic robustness and fairness. Furong is a recipient of the NSF CRII Award, the MLconf Industry Impact Research Award, the Adobe Faculty Research Award and three JP Morgan Faculty Research Awards. She is a Finalist of AI in Research - AI researcher of the year for Women in AI Awards North America 2022. She received her Ph.D. in electrical engineering and computer science from UC Irvine in 2016, after which she completed postdoctoral positions at Microsoft Research NYC.
Host: Yan Liu
Location: Seeley Wintersmith Mudd Memorial Hall (of Philosophy) (MHP) - 101
Audiences: Everyone Is Invited
Contact: Department of Computer Science
-
CS Colloquium: Keith Burghardt (USC ISI) - Utilizing Data Analysis To Reduce AI Biases
Wed, Oct 26, 2022 @ 02:00 PM - 03:00 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Keith Burghardt, USC ISI
Talk Title: Utilizing Data Analysis To Reduce AI Biases
Abstract: Biases are erroneous assumptions about data that can lead artificial intelligence (AI) systems to discriminate, policy makers to make harmful decisions, and data scientists to make conclusions that contradict reality. Biases, however, are often challenging to find or remove because they can be subtle and deeply embedded within data. In this talk, I will discuss how data can inadvertently create biases and present my research that aims to reduce them. I will first show how data can enhance biases, including how computer-human interactions can drive algorithmic ranking systems to erroneous conclusions and how anti-vaccine sentiment and hate speech can become prevalent on social media, which can lead to stereotypes embedded in AI language models. I will then discuss methods that reduce biases by utilizing data analysis. I will show how these methods can improve AI fairness and help researchers better understand how large systems, such as institutions and cities, evolve in time. I will conclude the talk by laying out how this work is a step towards the wider goal of AI risk minimization.
This lecture satisfies requirements for CSCI 591: Research Colloquium
Join Zoom Meeting
https://usc.zoom.us/j/93464447234?pwd=ZHlKeFlJVTBHWmhoTS9NRVBqTVV5QT09
Meeting ID: 934 6444 7234
Passcode: 475457
Biography: Burghardt is a Computer Scientist at the USC Information Sciences Institute who specializes in complex science, geospatial analysis, and reducing biases with data analysis. He has papers in journals such as NPJ Computational Materials and Communications Physics, and in conferences, such as ICWSM, ASONAM, and CSCW. Burghardt has been a PI in grants from Amazon and ISI, co-PI in grants from DARPA, and co-organized the Inclusive and Fair Speech Technologies special session at the INTERSPEECH 2022 Conference. Burghardt received a PhD and BS (Magna Cum Laude with High Honors) in Physics at the University of Maryland in 2016 and 2012, respectively.
Host: Vatsal Sharan
Audiences: Everyone Is Invited
Contact: Cherie Carter
-
CS Colloquium: Matteo Sesia (USC Marshall School of Business) - Conformal inference for uncertainty-aware classification
Thu, Oct 27, 2022 @ 03:30 PM - 04:50 PM
Thomas Lord Department of Computer Science
Conferences, Lectures, & Seminars
Speaker: Matteo Sesia, USC Marshall School of Business
Talk Title: Conformal inference for uncertainty-aware classification
Series: Computer Science Colloquium
Abstract: Complex machine learning classifiers, including deep neural networks, are sometimes able to achieve very high predictive accuracy, but they are not designed to realistically capture uncertainty or to estimate reliable probabilities. In fact, these models are often overconfident, and this issue can make it challenging for practitioners to accept the use of machine learning algorithms in delicate real-world applications. This talk will describe recent advances in the field of conformal inference which allow us to address the overconfidence of machine learning classifiers. First, this talk will present a powerful and statistically principled methodology for assessing the uncertainty of predictions computed by any pre-trained classification model, in such a way as to account for possible heterogeneity in the levels of uncertainty affecting different individual data points. Then, building upon the previous results, this talk will present a novel methodology for training deep neural networks in such a way as to learn multi-class classification models that are less prone to overconfidence, ultimately leading to even more reliable uncertainty-aware predictions.
Prof. Sesia will give his talk in person at RTH 115 and we will also host the talk over Zoom.
Join Zoom Meeting
https://usc.zoom.us/j/92821217575
Meeting ID: 928 2121 7575
This lecture satisfies requirements for CSCI 591: Research Colloquium.
Biography: Matteo Sesia is an assistant professor in the department of Data Sciences and Operation at the USC Marshall School of Business.
Matteo joined USC Marshall in 2020, immediately after earning a PhD in Statistics from Stanford University, where he was advised by Emmanuel Candes. Matteo's research primarily focuses on developing novel methodology for model-free statistical inference with big data, and on developing statistically principled algorithms for uncertainty-aware machine learning.
Host: Yan Liu
Webcast: https://usc.zoom.us/j/92821217575Location: Ronald Tutor Hall of Engineering (RTH) - 115
WebCast Link: https://usc.zoom.us/j/92821217575
Audiences: Everyone Is Invited
Contact: Department of Computer Science
