Sijie Xiong – Projects

The following listed all the projects I've done or I'm currently working on at Rutgers.

Research Projects

  • PERMIT: Privacy-Enabled Resource Management for IoT Networks (06/16 - present)

  • (Advisors: Prof. Anand Sarwate and Prof. Narayan Mandayam)

    - Studies how privacy, utility, and bandwidth affect each other in networked data collection and information processing systems.

  • Jana: Ensuring Secure, Private and Flexible Data Access (01/16 - 05/16)

  • (Advisors: Prof. Anand Sarwate, Prof. Desheng Zhang and Prof. Rebecca Wright)

    - Currently exploring the correlation between multi-modal geo-spatial data.

    - Poster demonstration on Rutgers ECE research day, 2017.

    - Developed an open-source Python library dp-stats for commonly used statistics and machine learning algorithms with differential privacy.

    - Functions included: Mean, Variance, Histogram, Principal Component Analysis, Support Vector Machines, Logistic Regression.

    - iPython notebook tutorials for each function.

    - Poster demonstration on Rutgers ECE research day, 2017.

  • Privacy-preserving learning for distributed data (06/15 - 12/15)

  • (Advisor: Prof. Anand Sarwate)

    - Designed a randomized requantization scheme which guarantees local differential privacy.

    - Looked at the tradeoff between privacy level, compression ratio and data utility.

    - Applied the randomized quantizer on synthetic data and showed its superiority than a baseline method when computing the average/aggregate statistics.

    - The project led to this publication.

Course Projects