Xinyu Liu

Experience

• 

  Singapore University of Technology and Design (SUTD)

  Jan 2017 - Dec 2017

  Teaching assistant for:• Advanced Mathematics II (Linear Algebra, Multivariate Calculus)• Modelling the Systems World (Optimization, Ordinary Differential Equations, System Modeling)• Data and Business Analytics (Data Visualization and Analytics with R, SQL, QGIS, Excel) Research title: Using optimization models to study responses to mass casualty events.• Explored the potential of graphical optimization models in visualizing and solving constrained resource allocation in the medical networks after bioterrorism attacks.• Investigated the formulation of complex post-bioterrorism logistic systems, revised the modeling using a graphical optimization tool and conducted analysis of simulated output.

  • Undergraduate Teaching Assistant

    Sept 2016 - Dec 2017

  • Undergraduate Research Assistant

    Jan 2017 - Apr 2017
• 

  SAP

  May 2017 - Aug 2017

  Data Science Research Intern

  Research Intern in Natural Language Processing and applications of Deep Neural Networks• Proposed a revised Sequence-to-Sequence Long Short-Term Memory network for generative Question-Answering system, enforcing semantic matching by separating the training of encoder and decoder while compensating the weakened syntactic learning through a pretraining scheme. • Experimented on large-scale public data sets of paired texts and demonstrated 80% accuracy out of 10,000 candidates on sematic matching component.• Performed validation on API (Application Programming Interface) products through experiments and machine learning analogies; created and communicated working solutions for client projects using analytics and visualization tools including Python and Microsoft Office. • Collaborated in a multifunctional team under the Agile development cycle. Show less

• 

  A*STAR - Agency for Science, Technology and Research

  Jan 2018 - Aug 2018

  Student Researcher

  • Designed a novel framework to derive semantic representation of human behaviors from videos using computer vision and natural language processing techniques. • Generated human-readable machine understanding of video content through a recurrent autoencoder; obtained state-of-the-art classification accuracies on behavior analysis subtasks; developed a use case on automatic content-based video tag creation; and nominated for SUTD Capstone (Final Year Project) Award.

• 

  Amazon

  May 2020 - Aug 2020

  Research Scientist Intern

  • Designed and validated an advanced fulfilling process, building upon the multi-echelon system. • Modeled and formulated a mixed integer linear program, implemented solution algorithms in Python, queried input data with SQL and produced numerical results demonstrating a maximum of 70% reduction in shipping delays at low or no additional cost.• Provided entitlement calculation and policy suggestions to the management and business teams, and successfully advocated for multiple prototypes of the proposed process. Show less

• 

  Georgia Institute of Technology

  May 2022 - Jul 2024

  ISyE Outstanding GTA for 2021Graduate Teaching Assistant for:• ISyE 3044 Simulation Analysis and Design• ISyE 3232 Stochastic Manufacturing and Service Systems• ISyE 4803 Network Science and Analysis• ISyE 4800/4106 Senior Design • ISyE Mission Possible Summer Camp• Seth Bonder Camp in Computational and Data Science Instructor-of-record for an undergraduate IE core course (class size of 31 students).Topics covered: Stochastic inventory models, Queueing theory, Discrete- and continuous-time Markov chains, and Poisson Process.

  • Graduate Teaching Assistant

    Aug 2018 - Jul 2024

  • Graduate Student Instructor

    May 2022 - Aug 2022
• 

  University of Michigan

  Aug 2024 - now

  Schmidt AI in Science Postdoctoral Fellow