Xiaomeng Ju

Xiaomeng Ju

PhD candidate in Statistics

Universit of British Columbia

Biography

I am a PhD student in Department of Statistics at the University of British Columbia under the supervision of Matías Salibián-Barrera. My research interests primarily lie in the area of computational statistics, with a special focus on robust statistics and functional data analysis. Previously, I received an MA in Statistics from University of Michigan-Ann Arbor and BSc in Statistics from Renmin University of China.

Interests
  • Computational Statistics
  • Robust Statistics
  • Functional Data Analysis
  • Machine learning
Education
  • PhD in Statistics, 2022 (Expected)

    University of British Columbia

  • MS in Statistics, 2015

    University of Michigan-Ann Arbor

  • BSc in Statistics, 2013

    Renmin University of China

Experience

 
 
 
 
 
Research and teaching assistant
University of British Columbia
Sep 2015 – Present Vancouver,BC
  • Worked on PhD thesis titled “Boosting for regression problems with complex data” under the supervision of Matías Salibián-Barrera. Developed gradient boosting algorithms for inhomogeneous data, functional data, and a combination of both.
  • Worked on the summer research project “High-dimensional regression with instrumental variables”. Extended two-stage regression with instrumental variables to high dimensional data and examined the performance of post-LASSO with SNP data for phenotype prediction.
  • Teaching assistant:
    — labs for STAT 200, STAT 251, STAT 344 in the Department of Statistics
    — labs DSCI 571, DSCI 513 for Master of Data Science
    — tutorials for CPSC 540 in the Department of Computer Science.
 
 
 
 
 
Research intern (Mitacs Accelerate program)
1QB Information Technologies, Inc
Jan 2016 – Dec 2020 Vancouver, BC
  • Applied deep learning to screen molecules for finding drug candidates; fitted graph convolutional networks (referring to Deepchem) to predict docking scores with molecular data provided by Vancouver Prostate Center.
  • Explored the use of active learning to select the molecules to be docked for regression and classification tasks.
 
 
 
 
 
Short Term Consulting Services (STCS)
University of British Columbia
Sep 2016 – Mar 2018 Vancouver, BC
  • Served on the managing team from Sep 2017. Managed consulting requests; organized meetings to discuss incoming projects.
  • Selected projects: Medical students’ health research (Prof. Erica Frank, UBC). Modelling beggiatoa population at aquaculture sites in British Columbia (Mainstream Biological Consulting Inc)
 
 
 
 
 
Data scientist intern
Ford Motor Credit Company
May 2015 – Aug 2015 Dearborn, MI, US
  • Developed a two-way survival model to predict the renewal rate of car lease contracts, which took into account the seasonal effect, lifetime effect, and time-varying covariates.

In progress

Xiaomeng Ju, Matías Salibián-Barrera. RTFBoost: robust tree-based functional boosting (Expected 2022).

  • Initial results were presented at International Conference on Robust Statistics, 2021. R package is avaiable here.

Software

Devaloped R packages: RTFBoost (on CRAN) and RTFBoost.