Bixing (Arin) Yan

Student, New York University

Basic Information

  • Physics
  • Data Science
  • Level II Candidate

Contact Information

Skill

Links

Education

  • M.S. in Data Science, New York University

    Specialty Orientation: Natural Language Processing, Data Science in Finance

  • PhD in Physics, National University of Singapore

    Specialty Orientation: Data Analytics

Working Experiences

  • Research Associate Internship

    Shopee E-Commerce, Garena, Singapore

    • Webpage fraud detection (anomaly detection, text mining) (Data size: 70GB)
    • SKU grouping (text mining, K-means, KNN) (Data size: 2.3GB) Link。

  • Summer Analyst Internship

    Industrial and Commercial Bank of China, Lanzhou

    • Helped to build the nonferrous future recommendation system based on ARIMA model.
    • Recommended nonferrous future based on fundamental analysis.

Research Experiences

  • Material Design with Neural Networks

    NanoCore Laboratory, National University of Singapore

    • Engineered the electronic and chemical features of oxides with PCA and xgboost
    • Designed a recurrent neural network to predict properties of correlated oxide. (Data size: 5GB)。

  • Polaron Modeling with Data Analytics

    Singapore National Research Foundation

    • Analyzed the transport data of anatase TiO2 with boosted tree, PCA, linear regression and K-means.
    • Modeled the evolution of polaron in anatase TiO2 based on the extracted features.
    • Interpreted the hot carrier dynamics through multi-exponential fitting. (Data size: 4.2GB)

  • Summer Research Internship

    Institute for Interdisciplinary Information Sciences, Tsinghua University

    • Simulated the spread of H7N9 avian influenza on small-world and scale-free networks.
    • Modelled the recession of H7N9 with modified logistic model.

Competition Experiences

  • Zestimate: Zillow’s Home Value Prediction

    Kaggle Competition

    • Preprocessed data with PCA (feature extraction), random forest (fix missing values), combination / exponentiation (create new features) and K-means (cluster the location).
    • Built a neural network with 5 hidden layers to predict estate price, with the mean absolute error of 0.6434 and ranked at 5% in 3800 teams. (Data size: 1.4GB)

Publications

  • Journal Pubulications

    • “Electron transport and visible light absorption in a plasmonic photocatalyst based on strontium niobate”. Nature Communication 8 (2017) 15070 Link
    • “Tunable and low-loss correlated plasmons in Mott-like insulating oxides.” Nature Communication 8 (2017) 15271 Link
    • “Anatase TiO2-a model system for large polaron transport.” American Chemical Society, Applied Materials and Interfaces 8b11643 (2018) Link
    • “Large polaron evolution with temperature and carrier density in anatase TiO2.” Bulletin of the American Physical Society: S17. 00004 (2018) Link
    • “Large polaron evolution in anatase TiO2 due to carrier and temperature dependence of electron-phonon coupling.” arXiv:1711.04107 (2018) Link

  • Conference Presentations

    • “Polaron Evolution in Anatase TiO2” (Talk) in American Physics Society, Mar 2018
    • “Correlated Metals as Transparent Conductive Oxide” in Oxide Electronics on Silicon Beyond Moore Workshop, April 2017
    • “Screening of the e-ph Interaction in Anatase TiO2” in International Workshop on Oxide Electronics 23, October 2016
    • “Polaron Evolution in Anatase TiO2” in 4th International Conference on Asian Network for Natural and Unnatural Materials, June 2016