Graduate

The far-from-equilibrium dynamics of closed quantum systems has become a central topic in condensed matter physics, due to incredible experimental advances in cold atomic and other systems. Concepts of quantum information have taken center stage in this context, with entanglement, in particular, playing a central role in the emergence of thermodynamics. Predicting the behavior of these quantities, however, is notoriously hard as most existing analytical and numerical tools, designed for systems in equilibrium, do not carry over to the dynamical setting. In my talk I will discuss how studying a set solvable minimal models has allowed us to partially conquer this problem and uncover universal features of quantum dynamics. In particular, I will describe some simple hydrodynamic models of how quantum information spreads in time, and discuss a recent prediction for entanglement growth that is particularly relevant for cold atom experiments.

This presentation overviews fluorescence lifetime spectroscopy and imaging techniques for label-free in vivo characterization of biological tissues. Emphasis is placed on recently developed devices and methods enabling real-time characterization and diagnosis of diseased tissues during clinical interventions. I will present studies conducted in animal models and human patients demonstrating the ability of these techniques to provide rapid in-situ evaluation of tissue biochemistry and their potential to guide surgical and intravascular procedures. Current results demonstrate that intrinsic fluorescence can provide useful contrast for the diagnosis of vulnerable atherosclerotic plaques, intraoperative delineation of brain tumors and head and neck tumors. Finally, I will present results from the first-in-human study that shows the potential of a multispectral fluorescence lifetime method for image-guided augmented reality in trans-oral robotic surgery (TORS).

The trend in neural recording capabilities is clear: we can record orders of magnitude more neurons now than we could only a few years ago, and technological advances do not seem to be slowing. Coupled with rich behavioral measurements, genetic sequencing, and connectomics, these datasets offer unprecedented opportunities to learn how neural circuits function. But they also pose serious modeling and algorithmic challenges. How do we develop probabilistic models for such heterogeneous data? How do we design models that are flexible enough to capture complex spatial and temporal patterns, yet interpretable enough to provide new insight? How do we construct algorithms to efficiently and reliably fit these models? I will present some of our recent work on recurrent switching linear dynamical systems and corresponding Bayesian inference algorithms that aim to overcome these challenges, and I will show how these methods can help us gain insight into complex neural and behavioral data.

In every corner of machine learning and statistics, there is a need for estimators that work not just in an idealized model but even when their assumptions are violated. Unfortunately, in high-dimensions, being provably robust and efficiently computable are often at odds with each other.

In this talk, we give the first efficient algorithm for estimating the parameters of a highdimensional Gaussian which is able to tolerate a constant fraction of corruptions that is independent of the dimension. Prior to our work, all known estimators either needed time exponential in the dimension to compute, or could tolerate only an inverse polynomial fraction of corruptions. Not only does our algorithm bridge the gap between robustness and algorithms, it turns out to be highly practical in a variety of settings.

In this talk, we'll focus on a class of resource allocation algorithms for communication networks: if a node of this network has L requests to transmit and is idle, it tries to access the channel at a rate proportional to log(1 + L). We'll study a simple stochastic model for such an algorithm in the case of a star network, in which J nodes can transmit simultaneously but interfere with a central node 0 in such a way that node 0 cannot transmit when one of the other nodes does. In contrast with the case where the probability of accessing the channel is proportional to L, as the total number of pending requests tends to infinity several timescales interact in a fine way to determine the asymptotic behaviour of the system. In particular, the numbers of pending requests at every node can evolve on very different timescales and have very different orders of magnitude. This is joint work with Philippe Robert (INRIA Paris).

Since the very beginning, X (formerly Google X) has had a single mission: to invent and launch "moonshot" technologies — like energy kites, balloon-powered internet, all-electric deliver drones, and self-driving cars — that we think could make the world a radically better place. Our "Rapid Evaluation" team, led by Phil Watson, is responsible for identifying, investigating, and evaluating these potential moonshots. In this talk, Phil will share what he's learned about finding and shaping promising new moonshots along their journey from outlandish, high-risk ideas into projects that just might have a shot at making it into the real world — and making a difference in the big picture challenges facing society.

What makes a successful engineering team? Why makes a fun engineering team? How can the two be combined? Following the adventures (and misadventures) of the the Stanford team competing in the University Rover Challenge, the talk will cover how to go about starting a multi person design project, how testing can be approached and we delve into some of the technical details of the two robots.

From subsea fiber cables to short-reach switch interconnects, opto-electronics is a key technology for hyperscale data center networks. As performance requirements increase, photonics moves deeper into the network replacing copper for shorter distances. The next move for photonics is to distances of less than 3m for in-rack applications. This talk will describe how the scale of data-bandwidth growth has challenged what is possible with traditional networks and where the next opportunities for innovation lie.

The Office of the Vice Provost for Graduate Education invites you to the Autumn quarter Diversity Works workshop, Communicating with Mastery presented by JD Schramm.

Please join us for this lunch discussion at the Denning House. This workshop is open to all Stanford graduate students.

Learn more and RSVP HERE by Sunday, October 20th.

Join our Diversity Works discussion with JD Schramm, director of the King Global Leadership Program at Knight-Hennessy and lecturer at the Graduate School of Business since 2007.

Many who've felt marginalized due to (race, gender, sexual orientation, etc.) struggle to communicate with strength when given the opportunity. In this engaging and interactive workshop JD Schramm will guide participants through a framework he's used with leaders in (and beyond) the LGBTQ community to overcome marginalization and communicate with strength and power.

About Diversity Works

Come together with academic leaders to discuss issues related to diversity and excellence in higher education at Diversity Works.

Open to enrolled graduate students at any stage in any discipline or degree program.

Please see the event detail page at Stanford US-Asia Technology Management Center.