Xinru Liu

Background: Identified similar business descriptions across diverse economic sectors using Hierarchy in NAICS Code.
Approach 1: Improved the baseline accuracy of semantic search for code prediction by 4% by developing a customized hierarchical similarity metric on MPNet embeddings through Breadth-first search (BFS) algorithm.
Approach 2: Boosted classification accuracy from 87% to 95% by engineering a hierarchical NLP classifier that constructs a layered representation linking business embeddings to a structured hierarchy of NAICS codes.

Redmond, WA 98052 · liu [dot] xinruuu [at] gmail [dot] com

Hello! I'm Xinru, a Data Scientist at Microsoft Azure Core. I hold a Ph.D. in Statistics from Boston University, where I was mentored by Yves Atchade. My research focuses on the theoretical and computational aspects of high-dimensional Bayesian statistics, with applications in sparse deep learning and Thompson Sampling. I obtained my B.A in Mathematics and Computer Science, minor in music performance from Wheaton College (MA). Besides work, I love exploring the arts, going to classical concert, playing the piano, and doing outdoor activities.

Publications & Preprints

Neural Thompson Sampling for Dynamic Pricing.
Joint work with Yves Atchade. Working paper.
On cyclical MCMC sampling.
Joint work with Liwei Wang, Aaron Matthew, and Yves Atchade. AISTATS, to appear, 2024.
Statistical perspective on unrolling models for inverse problems.
Joint work with Yves Atchade and Qiuyun Zhu. Submitted, 2023.
Multimodal Data Fusion in 3-D Printing Quality Prediction.

IEEE Sensors Letters, DOI 10.1109/LSENS.2018.2881475.

Besides Research

I start playing the classical music since I was five. I spend most of my leisure time practising the piano. Now I am exploring myself to more jazz and blues. Below are some videos of my performances for fun.