Hey there, my name is Nhu Ngoc, or simply Ngoc 👋 I'm a graduate student interested in machine learning, artificial intelligence, and developing data-intensive applications.

Erasmus Mundus Joint Master in Big Data Management and Analytics

I'm currently a master student pursing the Erasmus Mundus Joint Master Degree in Big Data Management and Analytics. I spent the first semester at Université Libre de Bruxelles (Belgium), and am currently spending my second semester at Universitat Politècnica de Catalunya (Spain). I will be spending my second year at Università degli Studi di Padova (Italy) to pursue the specialization in Statistics and Deep Learning for Data Analytics.

Bachelor of Science in Computer Science

In 2023, I graduated from New York University Abu Dhabi with a major in Computer Science and two minors in Applied Mathematics and Interactive Media. My undergraduate coursework and research experiences explored a wide range of topics in computer science but focused largely on machine learning and big data.

NYU Abu Dhabi's Center for Quantum and Topological Systems

Capstone project in Computer Science

NYU Tandon's Behavioral Urban Informatics, Logistics, and Transport Laboratory

NYU Abu Dhabi's Center for Global Sea Level Change

In early 2023, I worked at Wio Bank (Abu Dhabi, UAE) as an Application Security Intern.

From October 2023 to August 2024, I worked in R&D at LG Electronics Development Vietnam as a Software Engineer. My work focused on developing over-the-air firmware update solutions for in-vehicle infotainment (IVI) systems.

Erasmus Mundus Joint Master - Partner Country Scholarship

I was awared a full-ride scholarship by the European Commission to pursue the Erasmus Mundus Joint Master in Big Data Management and Analytics (BDMA). The scholarship is worth a total of €49,000, covering tuition fees, living expenses, and transportation costs for the two-year program.

New York University Founder's Day Award

This is an award for top-ranking graduating students at New York University.

• Hack Me If You Can: Aggregating AutoEncoders for Countering Persistent Access Threats Within Highly Imbalanced Data

  2024 Journal paper
  Read abstract

  Advanced Persistent Threats (APTs) are sophisticated, targeted cyberattacks designed to gain unauthorized access to systems and remain undetected for extended periods. To evade detection, APT cyberattacks deceive defense layers with breaches and exploits, thereby complicating exposure by traditional anomaly detection-based security methods. The challenge of detecting APTs with machine learning is compounded by the rarity of relevant datasets and the significant imbalance in the data, which makes the detection process highly burdensome. We present AE-APT, a deep learning-based tool for APT detection that features a family of AutoEncoder methods ranging from a basic one to a Transformer-based one. We evaluated our tool on a suite of provenance trace databases produced by the DARPA Transparent Computing program, where APT-like attacks constitute as little as 0.004% of the data. The datasets span multiple operating systems, including Android, Linux, BSD, and Windows, and cover two attack scenarios. The outcomes showed that AE-APT has significantly higher detection rates compared to its competitors, indicating superior performance in detecting and ranking anomalies.

  [published paper]

• NY Statewide Behavioral Equity Impact Decision Support Tool with Replica

  2023 Technical report
  Technical report I contributed to during my research assistanship at NYU Tandon's BUILT Lab.
  Read abstract

  A NY statewide model choice model is developed to deterministically fit heterogeneous coefficients for trips along each census block-group OD pair conducted by each population segment within a random-utility-consistent framework. The proposed approach is to use inverse optimization (IO) to derive coefficients for each OD pair times population segment as an agent. This is only possible with ubiquitous population data. We call this a group-level agent-based mixed logit (g-AMXL) model, which is an extension of the AMXL model proposed by Ren and Chow (2022). The significance of g-AMXL is as follows. First, g-AMXL takes OD level (instead of individual level) trip data as inputs, which is efficient in dealing with ubiquitous datasets containing millions of observations. Second, preference heterogeneities are based on non-parametric aggregation of coefficients per agent instead of having to assume a distributional fit. Third, since each agent’s representative utility function is fully specified, g-AMXL can be directly integrated into system design optimization models as constraints instead of dealing with simulation-based approaches required by mixed logit (MXL) models.

  [published report]

• A Vietnamese Named Entity Recognition System for COVID-19 Articles

  2022 Conference paper
  Research originally conducted for the class CSCI 480 Natural Language Processing at NYU Courant, Fall 2021. Paper later accepted at the 2022 IEEE MIT Undergraduate Research Technology Conference, Massachussets, United States. Due to visa issues, I could not present my paper in-person and instead gave the presentation via Zoom.
  DOI: 10.1109/URTC56832.2022.10002170.
  Read abstract

  This paper presents a named entity recognition system for the specific domain of Vietnamese COVID-19 news articles. By incorporating manually selected and domain-specific features into a simple deep learning architecture, the system can identify a wide range of custom named entities relevant in the context of COVID-19 and future epidemics. Using high-dimensional embedding vectors in combination with part-of-speech tags and additional features, the system achieves an F score of about 90.41%, surpassing or coming close to results by other models that are more complicated or pre-trained and fine-tuned.

  [published paper] [pre-recorded presentation]

• qSa'id (كيو-ساعِد) - Quantum-ML-Assisted Diagnostic Treatment Access Platform for Autism

  2022 Hackathon
  A two-fold project including a hybrid quantum - classical machine learning screening tool and an approach to the quadratic unconstrained binary optimization (QUBO) problem formulated as optimizing placements of specialized treatment centers in a healthcare system. Created for NYU Abu Dhabi Hackathon for Social Good in the Arab World 2022 - Focusing on Quantum Computing.
  [repo] [press coverage]

I come from Quảng Trị Province in Central Vietnam. I also spent three years in Hồ Chí Minh City during which I attended Lê Hồng Phong High School for the Gifted. Since undergraduate, I have lived, studied, and conducted research in many cities around the world, including Abu Dhabi, New York, Brussels, and Barcelona.

I dabble in the arts and generally anything and everything that fascinates me. I enjoy playing the flute, taking amateur photographs, creating digital arts, editing videos, and learning languages.

Unless you speak fluent Vietnamese, it's highly likely that you won't be able to pronounce my name correctly. I'm giving you the blanket approval to pronounce it /knock/ :)

◄ Featured photo: Sweihan, Abu Dhabi. Taken during a field trip in January 2020.