Carnegie Mellon University

Gauri Joshi

Gauri Joshi

Assistant Professor, Electrical and Computer Engineering

  • 4105 Collaborative Innovation Center
  • 412-268-1186
Address 5000 Forbes Avenue
Pittsburgh, PA 15213

Bio

Gauri Joshi is an assistant professor in the ECE department at Carnegie Mellon University since September 2017. Prior to that, she worked as a Research Staff Member at IBM T. J. Watson Research Center. Gauri received her Ph.D. from MIT EECS in June 2016 and received a B.Tech and M. Tech in Electrical Engineering from the Indian Institute of Technology (IIT) Bombay in 2010. Her awards and honors include the IBM Faculty Award (2018), Best Thesis Prize in Computer science at MIT (2012), Institute Gold Medal of IIT Bombay (2010), and the Claude Shannon Research Assistantship (2015-16).

Education

Ph.D., 2016
MIT

B. Tech, M. Tech., 2010
IIT Bombay

Research

Professor Joshi is interested in performance analysis and optimization of computing systems using a broad range of tools from probability, coding theory, and machine learning. Examples of current research themes are described below.

Efficient Redundancy in Cloud Systems

Cloud services need to ensure fast and seamless service to users. However the inherent randomness in response time of individual servers may cause large and unpredictable delays in serving users. A simple idea to reduce delay is to launch replicas of a task on multiple servers and wait for the earliest copy to finish. We seek a fundamental understanding of when such redundancy can outweigh the cost of additional resources. This research opens many interesting problems at the interface of coding and queueing theory.

Infrastructure for Distributed Machine Learning

The immense amount of data required to train state-of-the-art neural network models calls for a distributed infrastructure to process the data in parallel. The speed-up achieved by parallelizing is impeded by the time taken to synchronize all learners and ensure that they have up-to-date model parameters. A solution often used in practice is to simply run asynchronous model training, while running the risk of learners working with stale parameters. We aim to understand how these two factors: synchronization delays and parameter staleness affect the speed of convergence of the underlying algorithm.

Keywords

  • Performance modeling and analysis
  • Distributed storage and computing
  • Machine learning
  • Information theory

Related news

Tuesday, April 30, 2019

First round of Secure and Private IoT Initiative funded projects announced

CyLab’s Secure and Private IoT Initiative (IoT@CyLab) has broken ground as the first round of funded proposals have been announced. Twelve selected projects will be funded for one year, and results will be presented at the IoT@CyLab annual summit next year.
Thursday, April 11, 2019

Seed grants for energy research

The seventh round of the Wilton E. Scott Institute for Energy Innovation’s Seed Grants for Energy Research will provide funding to eight projects. These projects explore key energy topics including emerging information technology and advances in high-performance materials and natural gas solutions.
Monday, October 15, 2018

Optimizing computing systems

Joshi, assistant professor of Electrical and Computer Engineering (ECE), has been named a recipient of a 2018 IBM Faculty Award for her research in distributed machine learning.
Monday, July 23, 2018

Multi-armed bandits: using algorithms for sequential decision-making

ECE Assistant Professor Gauri Joshi is investigating multi-armed bandit algorithms to make data-driven inference and decision-making faster and more efficient. She recently received the Berkman Faculty Development Award to aid her research and acquire real-world data sets to validate her findings.
Wednesday, February 08, 2017

Meet the new ECE faculty members

As a new semester begins, we welcome four new faculty members to ECE. Meet Giulia Fanti, Carlee Joe-Wong, Gauri Joshi, and Bryan Parno.