Ph.D. student

Lehigh University

Biography

Yutong is a Ph.D. candidate in the Department of Industrial and Systems Engineering at Lehigh University, working under the supervision of Professor Daniel P. Robinson. He also has the fortune to work closely with Professor Lichao Sun.

His current research focuses on 1) designing, analyzing and implementing algorithms for large scale non-convex non-smooth optimization problems arisen in machine learning and federated learning; 2) making machine learning and deep learning algorithms secure, private, and robust.

Interests

  • Nonsmooth Optimizaiton with Structured Sparsity
  • Optimization in Machine Learning
  • Adversarial Machine Learning
  • Federated Learning

Education

  • M.S. in Statistics, 2019

    University of Illinois at Urbana-Champaign

  • B.S. in Statistics (with honors), 2017

    Sichuan University

Experience

 
 
 
 
 

Data Scientists Intern

Bud Analytic Lab of Anheuser-Busch InBev

Jan 2018 – May 2019 Urbana, IL
Leverage the power of the machine learning algorithms and domain knowledge to support global agronomist teams manage Barley in a smart way.

Publications

Quickly discover relevant content by filtering publications.

Local-Global Knowledge Distillation in Heterogeneous Federated Learning with Non-IID Data

Federated learning enables multiple clients to collaboratively learn a global model by periodically aggregating the clients’ models without transferring the local data. However, due to the …
Dezhong Yao, Wanning Pan, Yutong Dai, Yao Wan, Xiaofeng Ding, Hai Jin, Zheng Xu, Lichao Sun
Preprint

A Subspace Acceleration Method for Minimization Involving a Group Sparsity-Inducing Regularizer

We consider the problem of minimizing an objective function that is the sum of a convex function and a group sparsity-inducing regularizer. Problems that integrate such regularizers arise in modern …
Frank E. Curtis, Yutong Dai, Daniel P. Robinson
Preprint

Synchronous Parallel Block Coordinate Descent Method for Nonsmooth Convex Function Minimization

In this paper, we propose a synchronous parallel block coordinate descent algorithm(PSUM) for minimizing a composite function, which consists of a smooth convex function plus a non-smooth but …
Yutong Dai, Yang Weng
PDF

Projects

*
All Deep Learning Data Analysis

Real Estate Market Data Analysis

Develope data products to help Airbnb hosts to determine listing prices.

Show and Tell: A Neural Image Caption Generator

Feed in an image, AI will generate the caption for you!

Variational Gaussian Mixtures

Clustering under the variation inference framework.

Posts

Python Tricks Learned From Projects

Show all submodules I need to import a particular function formulate from a file in the directory <path-to-the-package>/coinor/dippy/examples/milp/milp_func. It’s clear that I need to import it from the submodule coinor.dippy. But how to do it exactly?
Last updated on Jan 17, 2021 python

中国传统计时

中国传统计时单位 古时一天分12个时辰，采用地支作为时辰名称，
Last updated on Aug 19, 2019 visualization
See all posts

Tags

Block Cooridinate Descent CNN Convex Optimization data heterogeneity data-analysis Deep Learning federated learning knowledge distillation LSTM proximal gradient python RandomForest second-order method subspace acceleration support identification visualization

Contact