Headshot Stock Photo

Office Hours: TBD

Classes

  • ECE 2300: Digital Logic Design
  • ECE 4300: Computer Architecture
  • EGR 4810: Project Design Principles and Applications
  • baquon at cpp.edu

    Fun fact: I own a coral cactus!
    Head to Student Resources for available stuff to use!


  • About


    I am an assistant professor in the Department of Electrical and Computer Engineering (ECE) at California State Polytechnic University, Pomona (Cal Poly, Pomona). Before joining Cal Poly Pomona, I spent a year as an adjunct professor at California State University, San Bernardino (CSUSB) teaching object oriented programming in C++, Digital Logic, and Platform Computing.

    I consider myself as a result of the public educational system. I earned my BS in computer engineering from CSUSB (2017) and graduated with my MS (2018) and PhD (2023) in computer engineering from University of California, Irvine (UCI).

    I have explored the entrepreneurial side of research and was a Microsoft Imagine Cup World Finalist (Education category) and extended the idea to NSFs Innovation Corps program at UCI.


    Research


    My research focus has evolved from concept drift and recommendation systems into multimodal ensembling, embodied cognitive learning, and data-centric AI.


    Publications

    B.A. Quon, J.L. Gaudiot

    Adaptive Aggregated Drift Detector,

    2023

    Data-centric Machine Learning Research (DMLR) Workshop at ICML 2023

    Hybrid Event - Honolulu, Hawai'i July 23 - 29, 2023

    There needs to be an adaptive approach that combines both performance and distribution based concept drift detectors in order to harness the benefits of unlabeled data and the ability to detect varying types of drifts. This paper proposes Adaptive Aggregated Drift Detector (A2D2), which consists of a suite of performance and data distribution based detectors that can adaptively select detectors based on rankings of least cost. The notable contribution is that it enables an ecosytem to not only adaptively combat drift, but to also expand the information learned across a suite of detectors.

    Read More

    B.A. Quon, J.L. Gaudiot

    Collaborative Concept Drift Detection,

    2023

    First Tiny Papers Track at ICLR 2023

    Hybrid Event - Kigali, Rwanda May 5, 2023

    Collaborative Concept Drift Detection (C2D2) combines Fast Correlated Based Filtering (FCBF) and Singular Value Decomposition (SVD) to detect concept drifts in 5 synthetic datasets. We compare our results against 6 diveregence tests and introduce Performance Gain Update Cost Ratio (PGUCR). Post-hoc Tukey HSD test confirmed that C2D2 outperformed the other tests in terms of PGUCR. Much of C2D2’s improvement is based on its conservative signals for updates.

    Read More

    B.A. Quon, J.L. Gaudiot

    Concept drift detection for distributed multi-model machine learning systems,

    2022

    2022 IEEE 46th Annual Computer Software and Applications Conference (COMPSAC)

    Virtual Event June 27 - July 1, 2022

    Many works focus on optimizing machine learning models during their training phase, but fail to account how these models adapt into their model-serving phase once they are deployed into real world applications. In this phase models must process through streams of data that can evolve over time and distort the relationship between incoming data, causing concept drift. This paper proposes leveraging the advantages of emerging features stores in order to improve concept drift detection on unlabeled, dynamic data streams across multiple models. Firstly, we introduce Drift Detection on Distributed Datasets (QuaD), which combines classical drift detectors to make use oflabeled and unlabeled data, and create local context (i.e. per live model) and global context (i.e. across multiple models). Secondly, we propose using feature store entities, SHAP values, and Collaborative Filtering (CF) to augment unlabeled data across multiple models. To the best of our knowledge, QuaD is the first work that examines the collective behavior of concept drift across multiple models and discerns associations between models that may share a susceptibility in a dynamic setting. QuaD uses a combination of performance-based and data distribution-based drift detectors and CF to capture varying types of concept drifts for labeled and unlabeled data streams and is modeled around the data abstraction provided by emerging feature stores.

    Read More

    Student Resources

    TBD


    Developed By ©Afro Coded
    Michelle Chilcutt & Annese Everage