YUDI ZHANG

515-***-**** adkbai@r.postjobfree.com

EDUCATION

Iowa State University Aug. 2017 - Expected 2022

Ph.D. Statistics

Shandong University Aug. 2013 - Jun. 2017

B.S. Statistics

RESEARCH EXPERIENCE

Unsupervised Clustering on Uncertain Categorical Data

Generalized three kmodes algorithms (Lloyd’s, MacQueen’s and Hartigan and Wong’s algorithms) for clus- tering categorical data with per-coordinate, noise-free uncertainty information.

Proposed optimization methods for those three improved kmodes algorithms to make it more possible to reach the global optima given limited time.

Distributed (wrapped C source code) our algorithms as a R package CClust available on GitHub. Genome Decoding by an Innovative Hidden Markov Model

Proposed a cutting-edge time inhomogeneous Hidden Markov Model (HMM) to genotype and phase short sequencing data from alleotetraploid individuals.

Used multinomial logistic regression to model the relationship between the hidden space and observed data. Designed a regularization penalty to discourage the undesirable biases in the estimated transition matrices.

Introduced inhomogeneous variable length Markov chains to reduce the enormous hidden space of HMM. A Likelihood Based Method to Classify Polyploid Genotype Joint work with Roshan Kulkarni

Proposed screening methods and statistical tests for determining the number of unique haplotypes.

Developed the whole genome genotype calling program and the simulation program in C++. Semisupervised Clustering for Single-Cell RNA-Seq

Adopted autoencoder to denoise UMI count data which is multinomial modelled with PCR error and se- quencing error correction. Then conducted semisupervised generative clustering in the encoded latent space. COURSE PROJECT

Improved Transfer Learning for Image Classi cation

Embedded several pretrained models obtained from VGG16 into Adaboost for classifying image data, im- proved the accuracy to almost 0.95 for CIFAR10 data. PUBLICATION

K. S. Dorman, X. Peng, Y. Zhang. (2020+). Denoising Methods for Inferring Microbiome Community Content and Abundance. Book chapter for \Statistical Analysis of Microbiome Data" in the Frontiers in Probability and the Statistical Sciences series by Springer. (Accepted) WORKING EXPERIENCE

Statistical consultant, Iowa State University May 2019 - Present Teaching assistant, Iowa State University Aug. 2017 - May 2019 AWARDS

Iowa State University Vera David Fellowship Aug. 2018 Shandong University Scholarship Sept. 2014, 2015, 2016 TECHNICAL STRENGTHS

C/C++, R, Python, Pytorch, MATLAB, SQL, Spark, SAS