• The Privacy Loss Distribution and its Privacy Loss Class: The Central Limit Theorem in Differential Privacy and Other Insights

    Mon, Feb 3, 2020, 4:30 pm to 5:30 pm

    Abstract: Quantifying the privacy loss of a privacy-preserving mechanism on potentially sensitive data is a complex and well-researched topic; the de-facto standard for privacy measures are ε-differential privacy (DP) and its versatile relaxation (ε,δ)-approximate differential privacy (ADP). Recently, novel variants of (A)DP emerged and focused on giving tighter privacy bounds under continual observation.

  • Practical Oblivious Computation

    Tue, Feb 4, 2020, 4:30 pm to 5:30 pm


    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 can enable privacy-preserving data mining for sensitive data (e.g., genomic or financial data), and allow businesses and individuals to monetize their data without compromising their privacy.

  • Distributed Compression, the Information Bottleneck and Cloud Radio Access Networks: A Unified Information Theoretic View

    Tue, Feb 11, 2020, 4:30 pm to 5:30 pm

    Abstract: This talk focuses on connections between relatively recent notions and variants of the Information Bottleneck and classical information theoretic frameworks such as: Remote Source-Coding; Information Combining; Common Reconstruction; The Wyner-Ahlswede-Korner Problem; The Efficiency of Investment Information; CEO Source Coding under Log-Loss and others. We overview the upink Cloud Radio Access Networks (CRAN) with oblivious processing, which is an attractive model for future wireless systems

  • Streaming Analytics for the Future Grid

    Thu, Dec 12, 2019, 3:00 pm to 4:00 pm


    How to conduct real-time analytics of streaming measurement data in the power grid? This talk offers a dynamic systems approach to utilizing data of different time scale for improved monitoring of the grid cyber and physical security. This talk presents how to leverage synchrophasor data dimensionality reduction and Robust Principal Component Analysis for early anomaly detection, visualization, and localization. The underlying theme of the work suggests the importance of integrating data with dynamic physical models in the smart grid.

  • Recent Advances in Non-Convex Distributed Optimization and Learning

    Mon, Nov 18, 2019, 4:30 pm to 5:30 pm


    We consider a class of distributed non-convex optimization problems, in which a number of agents are connected by a communication network, and they collectively optimize a sum of (possibly non-convex and non-smooth) local objective functions. This type of problem has gained some recent popularities, especially in the application of distributed training of deep neural networks.


Subscribe to Seminars