Stanford Online

※本サイトに掲載されているチャンネル情報や動画情報はYouTube公式のAPIを使って取得・表示しています。

Videos

動画一覧

動画数:2484件

Stanford Seminar - Improving Computational Efficiency for Powered Descent Guidance

Stanford Seminar - Improving Computational Efficiency for Powered Descent Guidance

May 24, 2024 Richard Linares, MIT Improving Computational Efficiency for Powered Descent Guidance via Transformer-based Tight Constraint Prediction Future spacecraft and surface robotic missions require increasingly capable autonomy stacks for exploring challenging and unstructured domains and trajectory optimization will be a cornerstone of such autonomy stacks. However, the optimization solvers required remain too slow for use on resource constrained flight-grade computers. In this work, we present Transformer-based Powered Descent Guidance (T-PDG), a scalable algorithm for reducing the computational complexity of the direct optimization formulation of the spacecraft-powered descent guidance problem. T-PDG uses data from prior runs of trajectory optimization algorithms to train a transformer neural network, which accurately predicts the relationship between problem parameters and the globally optimal solution for the powered descent guidance problem. The solution is encoded as the set of tight constraints corresponding to the constrained minimum-cost trajectory and the optimal final landing time. By leveraging the attention mechanism of transformer neural networks, large sequences of time series data can be accurately predicted when given only the spacecraft state and landing site parameters. When applied to the real problem of Mars-powered descent guidance, T-PDG reduces the time for computing the 3 degrees of freedom fuel-optimal trajectory when compared to lossless convexification, improving solution times by up to an order of magnitude. A safe and optimal solution is guaranteed by including a feasibility check in T-PDG before returning the final trajectory. About the speaker: https://aeroastro.mit.edu/people/richard-linares/ More about the course can be found here: https://stanfordasl.github.io/robotics_seminar/ View the entire AA289 Stanford Robotics and Autonomous Systems Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMeercb-kvGLUrOq4HR6BZD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore #Stanford #Stanford Online
2024年06月22日
00:00:00 - 00:42:16
Stanford Seminar - How Can Privacy Exist in a Data-Driven World?

Stanford Seminar - How Can Privacy Exist in a Data-Driven World?

May 24, 2024 Blase Ur, University of Chicago Huge amounts of personal data underpin the algorithms that drive modern life. How can privacy exist in such a world, and what does privacy even mean in this context? In this talk, I will partially answer these questions by discussing how our group employs data-driven methods to help users understand how their data is collected and used. In particular, I will present tools we have developed both to provide transparency about online tracking and to help users engage with the personal data companies hold about them. To further contextualize the meaning of privacy, I will describe user studies investigating how privacy is perceived. I will conclude by describing our ongoing collaborations with artists to recenter privacy as a societal value. About the speaker: Blase Ur is an Associate Professor of Computer Science at the University of Chicago, where he researches computer security, privacy, human-computer interaction, and ethical AI. His lab, the UChicago SUPERgroup, uses data-driven methods to make complex computer systems more usable and to help users make better security and privacy decisions. He has received an NSF CAREER Award, Quantrell Award for Undergraduate Teaching, five best/distinguished paper awards, and five honorable mention paper awards. He has also received the Allen Newell Award for Research Excellence, SIGCHI Outstanding Dissertation Award, IEEE Cybersecurity Award for Practice, and a Fulbright scholarship to Hungary. He holds degrees from Carnegie Mellon University (PhD and MS) and Harvard University (AB). He also likes bicycles, photography, punk rock, and cacti/succulents. More about the course can be found here: https://hci.stanford.edu/seminar/ View the entire CS547 Stanford Human-Computer Interaction Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMyupDF2O00r19JsmolyXdD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore #Stanford #Stanford Online
2024年06月20日
00:00:00 - 01:00:30
Stanford AA222 / CS361 Engineering Design Optimization I Linear Constrained Optimization

Stanford AA222 / CS361 Engineering Design Optimization I Linear Constrained Optimization

April 25, 2024 Joshua Ott of Stanford University Learn more about the speaker: https://profiles.stanford.edu/joshua-ott This course covers the design of engineering systems within a formal optimization framework. This course covers the mathematical and algorithmic fundamentals of optimization, including derivative and derivative-free approaches for both linear and non-linear problems, with an emphasis on multidisciplinary design optimization. Topics will also include quantitative methodologies for addressing various challenges, such as accommodating multiple objectives, automating differentiation, handling uncertainty in evaluations, selecting design points for experimentation, and principled methods for optimization when evaluations are expensive. Applications range from the design of aircraft to automated vehicles. Visit the course website: https://aa222.stanford.edu/ Enroll in the course: https://online.stanford.edu/courses/aa222-engineering-design-optimization #Stanford #Stanford Online
2024年06月15日
00:00:00 - 01:19:31
Stanford Seminar - From the surface of Mars to the ocean of Enceladus

Stanford Seminar - From the surface of Mars to the ocean of Enceladus

May 17, 2024 Hiro Ono, NASA JPL From the surface of Mars to the ocean of Enceladus: EELS Robot to Spearhead a New One-Shot Exploration Paradigm with Risk-Aware Adaptive Autonomy NASA’s Perseverance rover, on its mission to find a sign of ancient Martian life that might have existed billions of years ago, has been enormously successful partially owing to its highly advanced autonomous driving capabilities. However, current Mars exploration requires ample environmental knowledge accumulated over decades and across multiple missions, resulting in slow progression towards exploring unvisited worlds beyond Mars. The EELS (Exobiology Extant Life Surveyor) robot, a snake-like robot designed for exploring extreme environments, aims to shift this exploration paradigm by utilizing versatile robotic hardware, mechanical flexibility, and intelligent, risk-aware autonomy. For the first time, this adaptive robot gives us the opportunity to explore environments currently out of reach. The ultimate mission of EELS would be exploring Saturn’s Enceladus geysers – searching within a subsurface ocean for extant alien life. We built hardware and software prototypes of EELS and successfully tested in a wide range of environment, including natural vertical holes on Athabasca Glacier in Canada. This talk will cover a broad range of topics related to autonomous robotic exploration of unknown planetary environments, including EELS, Mars rover autonomy, and risk-aware planning algorithms. About the speaker: https://www-robotics.jpl.nasa.gov/who-we-are/people/masahiro_ono/ More about the course can be found here: https://stanfordasl.github.io/robotics_seminar/ View the entire AA289 Stanford Robotics and Autonomous Systems Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMeercb-kvGLUrOq4HR6BZD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore #Stanford #Stanford Online
2024年06月14日
00:00:00 - 00:52:41
Stanford CS25: V4 I Hyung Won Chung of OpenAI

Stanford CS25: V4 I Hyung Won Chung of OpenAI

April 11, 2024 Speaker: Hyung Won Chung, OpenAI Shaping the Future of AI from the History of Transformer AI is developing at such an overwhelming pace that it is hard to keep up. Instead of spending all our energy catching up with the latest development, I argue that we should study the change itself. First step is to identify and understand the driving force behind the change. For AI, it is the exponentially cheaper compute and associated scaling. I will provide a highly-opinionated view on the early history of Transformer architectures, focusing on what motivated each development and how each became less relevant with more compute. This analysis will help us connect the past and present in a unified perspective, which in turn makes it more manageable to project where the field is heading. Slides here: https://docs.google.com/presentation/d/1u05yQQaw4QXLVYGLI6o3YoFHv6eC3YN8GvWD8JMumpE/edit#slide=id.g2885e521b53_0_0 0:00 Introduction 2:05 Identifying and understanding the dominant driving force behind AI. 15:18 Overview of Transformer architectures: encoder-decoder, encoder-only and decoder-only 23:29 Differences between encoder-decoder and decoder-only, and rationale for encoder-decoder’s additional structures from the perspective of scaling. About the speaker: Hyung Won Chung is a research scientist at OpenAI ChatGPT team. He has worked on various aspects of Large Language Models: pre-training, instruction fine-tuning, reinforcement learning with human feedback, reasoning, multilinguality, parallelism strategies, etc. Some of the notable work includes scaling Flan paper (Flan-T5, Flan-PaLM) and T5X, the training framework used to train the PaLM language model. Before OpenAI, he was at Google Brain and before that he received a PhD from MIT. More about the course can be found here: https://web.stanford.edu/class/cs25/ View the entire CS25 Transformers United playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rNiJRchCzutFw5ItR_Z27CM #Stanford #Stanford Online
2024年06月12日
00:00:00 - 00:36:31
Stanford Seminar - Replication strategies for more robust human simulation

Stanford Seminar - Replication strategies for more robust human simulation

May 17, 2024 Aaron Shaw, Northwestern University Increasingly, Large Language Models (LLMs) are used to simulate human behavior and social systems. However, despite rapidly growing scientific and commercial applications of LLMs along these lines, threats to the validity and robustness of such applications remain poorly understood and responses to these threats remain ad hoc. Replication strategies inspired by prior social science and statistical research offer insights into these challenges. This talk characterizes some key threats to LLM simulations of human behavior and explores two strategies---perturbation and iteration---to evaluate LLM simulations of human behavior in the context of social scientific replication. About the speaker: Aaron Shaw is Associate Professor of Communication Studies and Sociology (by courtesy) at Northwestern University and a Faculty Associate of the Berkman Klein Center for Internet and Society at Harvard University. He is a co-founder of the Community Data Science Collective. Around Northwestern, he is also affiliated with the Center for Human-Computer Interaction + Design (HCI+D), the Institute for Policy Research, and the Buffett Institute for Global Affairs. More about the course can be found here: https://hci.stanford.edu/seminar/ View the entire CS547 Stanford Human-Computer Interaction Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMyupDF2O00r19JsmolyXdD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore #Stanford #Stanford Online
2024年06月08日
00:00:00 - 00:57:06
Stanford CS25: V4 I Behind the Scenes of LLM Pre-training: StarCoder Use Case

Stanford CS25: V4 I Behind the Scenes of LLM Pre-training: StarCoder Use Case

May 23, 2024 Speaker: Loubna Ben Allal, Hugging Face As large language models (LLMs) become essential to many AI products, learning to pretrain and fine-tune them is now crucial. In this talk, we will explore the intricacies of training LLMs from scratch, including lessons on scaling laws and data curation. Then, we will study the StarCoder use case as an example of LLMs tailored for code, highlighting how their development differs from standard LLMs. Additionally, we will discuss important aspects of data governance and evaluation, crucial elements in today's conversations about LLMs and AI that are frequently overshadowed by the pre-training discussions. About the speaker: Loubna Ben Allal is a Machine Learning Engineer in the Science team at Hugging Face working on Large Language Models for code & Synthetic data generation. She is part of the core team behind the BigCode Project and has co-authored The Stack dataset and StarCoder models for code generation. Loubna holds Mathematics & Deep Learning Master's Degrees from Ecole des Mines de Nancy and ENS Paris Saclay. More about the course can be found here: https://web.stanford.edu/class/cs25/ View the entire CS25 Transformers United playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rNiJRchCzutFw5ItR_Z27CM #Stanford #Stanford Online
2024年06月08日
00:00:00 - 01:01:37
Stanford Seminar - When Design = Planning

Stanford Seminar - When Design = Planning

May 10, 2024 Cynthia Sung, UPenn Robot design is an inherently difficult process that requires balancing multiple different aspects: kinematics and geometry, materials and compliance, actuation, fabrication, control complexity, power, and more. Computational design systems aim to simplify this process by helping designers check whether their designs are feasible and interdependencies are satisfied. But what can we say about when a design that accomplishes a task even exists? Or what the simplest design that does a job is? In this talk, I will discuss recent work from my group in which we have discovered that, in some cases, design problems can be mapped to problems in robot planning, and that results derived in the planning space allow us to make formal statements about design feasibility. These ideas apply to systems as varied as traditional robot arms, dynamical quadrupeds, compliant manipulators, and modular truss structures. I will share examples from systems developed in my group and forecast forward on the implications of these results for future robot co-design. About the speaker: https://sung.seas.upenn.edu/people/sung/ More about the course can be found here: https://stanfordasl.github.io/robotics_seminar/ View the entire AA289 Stanford Robotics and Autonomous Systems Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMeercb-kvGLUrOq4HR6BZD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore #Stanford #Stanford Online
2024年06月07日
00:00:00 - 00:47:56
Demystifying Product Management: Your Questions, Expert Answers

Demystifying Product Management: Your Questions, Expert Answers

Learn everything you need to succeed as a product manager: https://online.stanford.edu/programs/product-management-program Whether you’re a seasoned product manager or looking to start your journey into the career, there’s a lot to know and even more to learn. Professor Mike Lepech hosted a Q&A with Anand Subramani, experienced PM and Stanford instructor, to explore the Stanford Online community's questions on product management. #Stanford #Stanford Online
2024年06月01日
00:00:00 - 00:26:31
Stanford Seminar - Online communities as model systems for commons governance

Stanford Seminar - Online communities as model systems for commons governance

May 10, 2024 Seth Frey, UC Davis The best citizens of a large-scale democracy are those who have built and broken several small ones to see how they work. By empowering people to build any kind of community together, the Internet has become a laboratory for self-governance experimentation. Groups who start online communities must overcome the challenges of recruiting finite resources around difficult common goals. Fortunately, they can draw on a growing range of support technologies, peer networks, and scholarship. With their transparency, the Internet's millions of online communities can be surveyed for insights into their design and functioning. Looking at three large platforms for small self-governing online communities, we will pose several questions of institutional processes at the population level, as drawn from the literatures on common-pool resource management and institutional analysis and design. About the speaker: Dr. Seth Frey is a computational social scientist who studies commons governance institutions and other complex social systems. He specializes in using online communities as model systems for emergent institutional and organizational phenomena. His expertise is in computational approaches to self-governance and the cognitive science of strategic behavior. He is an associate professor in Communication at the University of California Davis, an affiliate of the Ostrom Workshop at Indiana University, and a Research Director at Metagov. He was a behavioral economist at Disney Research in Walt Disney Imagineering, and a complex systems scholar at NECSI. Seth earned his Ph.D. in Cognitive Science and Informatics (complex systems) at Indiana University in 2013, and a B.A. in Cognitive Science from UC Berkeley. Seth's research has appeared in PNAS, Nature Scientific Reports, and Proceedings of the Royal Society. It has been covered in The New Yorker, New York Times Magazine, and TEDx. It has been funded by the NSF, NASA, and the Ford Foundation. More about the course can be found here: https://hci.stanford.edu/seminar/ View the entire CS547 Stanford Human-Computer Interaction Seminar playlist: https://www.youtube.com/playlist?list=PLoROMvodv4rMyupDF2O00r19JsmolyXdD ► Check out the entire catalog of courses and programs available through Stanford Online: https://online.stanford.edu/explore #Stanford #Stanford Online
2024年06月01日
00:00:00 - 00:58:00