University of Maryland
Host: Prof. Srini Devadas
Oblivious computation refers to the ability to compute on "encrypted" data, such that neither intermediate results nor the program's runtime behavior reveal anything about secret inputs. Oblivious computation carries great promise in various domains such as financial and genomic applications, and can enable businesses and individuals to monetize their data without compromising their privacy.
Two essential challenges lie in the way of making oblivious computation practical: 1) the efficiency of foundational cryptographic building blocks; and 2) ease-of-adoption by non-specialist programmers. In this talk, I will describe our efforts that combine algorithms, programming languages, and systems building to overcome these challenges.
I will first present a novel, binary-tree based paradigm for constructing Oblivious RAM (ORAM) schemes. ORAM is a generic and powerful primitive that is central to realizing oblivious computation based on either trusted hardware or cryptographic secure computation. Our new ORAM constructions not only solve a twenty-seven year open theoretical challenge, but also allow ORAM to evolve from a theoretical primitive to a practical building block. Specifically, we made it possible, for the first time, to implement ORAM atop secure processors as well as secure multi-party computation.
Next, I will present programming language techniques that not only accelerate oblivious computation, but also make it accessible to real-life programmers who are not security experts. Based on these algorithmic and programming language advances, we have built new platforms for oblivious computation and developed various demo applications such as common data structures, data mining, graph algorithms, and streaming algorithms. Evaluation results suggest that over a duration of three years, our work has led to four to five orders of magnitude performance improvement for moderately large data sizes, and has enabled a dramatic reduction in application development effort. Part of our framework is in the process of being publicly released via http://www.oblivm.com.
Elaine Shi is an Assistant Professor in the Department of Computer Science at the University of Maryland. Her research combines theory, programming languages, and systems techniques to design computing platforms that are efficient, easy to program, and provably secure. Elaine's research has been recognized with several awards, including an NSA Best Scientific Cybersecurity Paper Award, a UMD Invention of the Year Award, and an ACM CCS Best Student Paper Award. Elaine is the recipient of a Sloan Research Fellowship (2014), Google Faculty Research Awards (2013 and 2014), and winner of the IJCNN/Kaggle Social Network Challenge (2011). Elaine obtained her Ph.D. from Carnegie Mellon University. Prior to joining Maryland, she was a research scientist at the Palo Alto Research Center (PARC) and UC Berkeley.
Building 32 Map
A great idea does not guarantee great profits. If a company's R&D dollars are going to pay off in profitable products and technologies, it needs a strategy that not only makes markets, but also beats the competition. This program will present a depth of challenges that extend from R&D to manufacturing, engineering, project management, and new ventures, and provide an innovative and powerful approach to developing technologies and products that people want to buy. The program material will also explore ways to link those technologies and products with a company's business strategy.
Drawn from MIT Sloan School's top-ranked MBA curriculum, this groundbreaking program will provide a framework for understanding how technologies and markets evolve; how they are linked; how technologies differ across markets; and how new technologies get accepted. This program will enable participants to:
* Identify profitable projects for their research dollars and find out how to capture the value of those projects
* Build technical capabilities for products that create value for their customers
* Restructure their organizations to respond to market and technical dynamics
* Implement their strategies for maximum benefit
This program is essential for senior general and technical executives involved in developing, managing, or marketing technology or products, or with managing organizations that sell their products in rapidly changing markets. The program will be most beneficial for:
* Managers in technology-intensive organizations
* Marketing and business development executives in technology organizations
* R&D managers in any organization
Titles of past participants have included:
* Executive VP
* Head of R&D; Engineering; Manufacturing & IS
* VP of Marketing & New Venture Development
* Chief Technologist
* Corporate Planner; Strategists
This program is about how to analyze projects from a financial perspective. It will offer a unique opportunity for program and project managers to learn cost accounting-based project management practices and strategies for making smart project choices which justify outcomes and create value. Program material is drawn from our popular and highly-rated MBA courses on financial and managerial accounting and shows how managers can leverage cost analysis to better influence the outcomes of product development and project management.
The program will offer a series of interactive lectures, cases, and small group exercises that will help participants better understand:
* The language and mechanics of the accounting that goes on in complex organizations
* How to identify good results even though the accounting numbers look bad, and bad results when the accounting numbers look good
* Cost allocations, absorption costing, and transfer pricing, and their effect on reported performance
* Company’s internal metrics for evaluating management
This program has been developed for senior program and project managers from a wide range of consumer and business-to-business industries, including:
* Managers from engineering, manufacturing, IT and technology departments
* Directors of project management, product and business development, and R&D
* Chief project engineers
* Product design and process development engineers
* Key staff members with performance responsibility
This program is built around MIT's unique Distributed Leadership Model-a powerful, innovative approach to executive leadership that lies at the core of leadership development at MIT, and the result of an intensive, four-year research project at the MIT Leadership Center to identify more effective strategies for leading in a networked economy. Tested in diverse, real-world settings, the model allows managers to succeed as leaders by being flexible and adaptive in new and unexpected ways through the application of two key concepts:
* A 4 Capabilities Leadership Framework that makes it possible to harness, align, and leverage the leadership capabilities that exist throughout an organization, and
* X-Teams, a revolutionary approach to creating flexible, outwardly-focused project teams that enables managers to both keep current with shifts in markets, technologies, and competition, and accelerate the pace of innovation and change
Upon completion of this program, participants will gain an understanding of how to:
* Innovate and move quickly from generating ideas to executing and diffusing them throughout the organization
* Unlock crucial information, expertise, and new ways of working together, wherever these qualities reside within or outside the company
* Succeed in a competitive “flat world” of new organizational architectures; smart, orchestrated networks; and tiny firms that do not need huge capitalization to compete
* Make their organizations more agile, responsive, and creative
This program has been designed especially for senior general and technical executives whose organizations compete in an environment of rapidly changing markets, technologies, and cultures, including:
* Executive VPs
* Heads of R&D, Engineering, Manufacturing & IS
* Chief Technologists
* Corporate Planners and Strategists
* VPs of Marketing and New Venture Development
* Other senior managers with leadership responsibility
Seley Distinguished Professor of Management
Professor of Organization Studies
Faculty Director, MIT Leadership Center
Deborah Ancona's pioneering research explores how successful teams operate and the critical importance of managing both outside and inside the team's boundary. This research led directly to the concept of X-teams as a vehicle for driving innovation within large organizations and to the publication of her book, X-teams: How to Build Teams That Lead, Innovate, and Succeed (Harvard Business School Press, June 2007).
Ancona's work also focuses on the concept of distributed leadership and the development of research-based tools, practices, and teaching/coaching models that enable organizations to foster creative leadership at every level. This work was highlighted in the Harvard Business Review article, "In Praise of the Incomplete Leader" (February 2007).
Her studies of team performance have been published in the Administrative Science Quarterly, the Academy of Management Review, Organization Science, and the Sloan Management Review. Ancona's book, Managing for the Future: Organizational Behavior and Processes (South-Western College Publishing, 1996, 1999, 2005), centers on the skills and processes needed to succeed in today's diverse and changing organization. She has served as a consultant on leadership and innovation to premier companies, such as AT&T, BP, Credit Suisse First Boston, Hewlett-Packard, Merrill Lynch, News Corporation, and Vale.
Ancona holds a BA and an MS in psychology from the University of Pennsylvania and a PhD in management from Columbia University.
MIT general map location link
Professor Philippe Fargues
European University Institute
Sponsored by: Center for International Studies, Inter-University Committee on International Migration
Building 32 Map
The demand for wireless data is projected to continue to rapidly grow. Meeting this demand will require that wireless data services have access to additional RF spectrum. It is widely recognized that this in turn will require new approaches for spectrum management. In particular various mechanisms for sharing spectrum among multiple firms are attracting much interest. While such shared spectrum can allow for market expansion, its addition will also influence the competition among wireless service providers and greater sharing can increase the congestion seen by consumers of these services. In this talk we present several stylized models to provide insight into how different approaches to spectrum sharing may impact economic welfare. These are based on game theoretic models for price or quantity competition among firms with congestible resources, where here the congestion depends in part on how spectrum is shared.
Randall Berry joined Northwestern University in 2000, where he is currently a Professor in the Department of Electrical Engineering and Computer Science. He received the M.S. and PhD degrees in Electrical Engineering and Computer Science from MIT in 1996 and 2000, respectively. His undergraduate education was at the University of Missouri-Rolla, where he received the B.S. degree in Electrical Engineering in 1993. In 1998 he was on the technical staff at MIT Lincoln Laboratory in the Advanced Networks Group. Dr. Berry is the recipient of a 2003 CAREER award from the National Science Foundation. He is an IEEE Communications Society Distinguished Lecturer for 2013-14. He has served as an Editor for the IEEE Transactions on Wireless Communications from 2006 to 2009, and an Associate Editor for the IEEE Transactions on Information Theory from 2009 to 2011, in the area of communication networks. He has served on the program and organizing committees of numerous conferences including serving as the co-chair of the 2012 IEEE Communication Theory Workshop and a technical co-chair of 2010 IEEE ICC Wireless Networking Symposium. He is an IEEE Fellow.
Building 46 Map
Samuel Gershman, Ph.D.
A currently popular view of reinforcement learning is that the brain makes use of both model-based (goal-directed) and model-free (habitual) learning systems. I will discuss an alternative learning system, whose core is the successor representation (SR), which exhibits some of the operating characteristics of both model-based and model-free learning. The SR compactly encodes the manifold structure of the state transition function, and is closely related to spectral techniques for manifold learning. When applied to spatial environments, the SR learns a cognitive map that captures the underlying geometry of the state space. Such a map is consistent with the hippocampal representation of space: a variety of place field phenomena, such as changes in place fields induced by manipulations of environmental geometry and reward, arise naturally from the SR. Moreover, an eigendecomposition of the SR leads to a spatial representation resembling entorhinal grid cells, which may be useful for making the map robust to noise. I will describe recent behavioral experiments that provide evidence for the SR.