Songzhu Zheng

Songzhu Zheng

PhD Student of Statistics

Stony Brook University

Biography

I’m a statistics PhD student at Stony Brook University. My research focus on robust machine learning methods against label noise or data posioning attack. I’m also interested in the application of machine leanring method to solve financial problem.

Download my resumé.

Interests
  • Artificial Intelligence
  • Machine Learning
  • Statistical Inference
Education

  • PhD in Statistics, 2022 (expected)

    Stony Brook University

  • MA in Statistics, 2017

    Rice University

  • BS in Statistics, 2015

    Communication University of China

Publications

Quickly discover relevant content by filtering publications.
Songzhu Zheng, Yikai Zhang, Hubert Wagner, Mayank Goswami, Chao Chen (2021). Topological Detection of Trojaned Neural Networks. NeurIPS.

PDF

Yikai Zhang+, Songzhu Zheng+, Pengxiang Wu+, Mayank Goswami, Chao Chen (2021). (Spotlight, +equalcontribution) Learning with Feature Dependent Label Noise: a Progressive Approach. ICLR.

PDF Code

Pengxiang Wu+, Songzhu Zheng+, Mayank Goswami, Dimitris Metaxas, Chao Chen (2020). (+equal contribution) A Topological Filter for Learning with Label Noise. NeurIPS.

PDF Code

Songzhu Zheng, Pengxiang Wu, Aman Goswami, Mayank Goswami, Dimitris Metaxas, Chao Chen (2020). Error-Bounded Correction of Noisy Labels. ICML.

PDF Code Slides

Experiences

 
 
 
 
 
AI Researcher Intern
Morgan Stanley
Jun 2021 – Aug 2021 NYC

Responsibilities include:

  • Design DNN that identifies learnable data out of majority noisy dataset
  • Construct optimal MBS products with maximized market value with RL

Projects

*
All Machine Learning Other
Exploring the Relationship Between Cancer and Unhealthy Habits
Investigate the top risk factors that correlate with cancer cases using demographic data
Songzhu Zheng
Slides
Exploring the Relationship Between Cancer and Unhealthy Habits
Forecasting the Volatility of Stock Market with Bayes Stochastic Volatility Model
Garch model in Bayes frame work - an application of PyMc package
Songzhu Zheng
PDF
Forecasting the Volatility of Stock Market with Bayes Stochastic Volatility Model
Grammatical Facial Expression Recognition
Recognize the grammical facial expression using machine learning
Songzhu Zheng
Code Slides
Grammatical Facial Expression Recognition

Recent Posts

Build a R Package with Rcpp in 30 minutes
Rcpp is package with which you could write c++ code within R. You could simply write a inline C++ code in R like this: library(Rcpp) # Define a Rcpp function add_vec <- Rcpp::cppFunction( 'Rcpp::NumericVector add_vec( Rcpp::NumericVector a, Rcpp::NumericVector b){ return a+b; }' ) # See the result t(add_vec(c(1,2,3), c(2,3,4))) Or you could save your .
Songzhu Zheng
Last updated on Jan 15, 2021 4 min read Tool

Contact