ECE 559 - Statistical Approaches to Data Privacy

Spring 2014

Brief Outline

Collection of terabytes of data about individuals and the corresponding analytics are hallmarks of the modern information age. At times the data collection is enforced by government agencies, employers and the like. At other times, data is ``voluntarily" collected via economic incentives (frequent shopper cards, etc). On the other hand, the collected data is disseminated publicly (US Census data, for instance) or to third-party firms or via recommendation systems. This course studies statistical approaches to preserving individual privacy in databases and opinion collections. The course covers the wide range of technical developments in the area of differential privacy.

Detailed Course Outline

Location and Time: Monday and Wednesday, 10-11.20am, 1214 Siebel Center.


The primary audience for the course are graduate students with a sufficient background and mathematical maturity in probability and algorithms. A strong undergraduate course in probability (such as ECE 313) is essential. The mathematical maturity of handling a course such as ECE 534 or CS 473 or STAT 510 or equivalent is strongly recommended.

There is a take home midterm (35\%), a take home final (35\%) and a project with a written report component (30\%) in the course.

There is no text book for the course. A set of notes summarizing each lecture will be provided. Relevant papers from the literature will be provided as required reading prior to each lecture.

Teaching Staff

Instructor: Professor Pramod Viswanath
Office Hours: TBD
Phone: 244-8999
E-mail: pramodv at

Teaching Assistant: TBD