Student Led Events

Statistics & Data Science Student Journal Seminar

The SDS Student Seminars are opportunities for students to read a paper and present the material to the SDS student community. 

Statistics & Data Science Graduate students are encouraged to participate in the seminars, suggest topics or papers for discussion, or volunteer to present a paper or their own research.

Spring 2024 Schedule

Date/Time Location Presenter Title/Topic Additional Information
02/12/24 ENR2 S395 Cody Melcher Accelerating Stochastic Gradient Descent using Predictive Variance Reduction
03/15/24, 10am ENR2 S375 Jacob Maibach The Effect of SGD Batch Size on Autoencoder Learning Abstract
03/29/24 ENR2 S375 Zihan Zhu Bayesian Poisson Regression with Spatially Dependent Global-Local Shrinkage Prior Abstract

 

 

Fall 2023 Schedule

Date/Time Location Presenter Title/Topic Additional Information
09/18/2023, 2:30pm PAS 522 Shudong Sun "Neyman-Pearson Classification Algorithms and NP Receiver Operating Characteristics" 
Journal Paper 091823.pdf
10/09/2023, 2:30pm MATH 501 Jeffrey Mei Reconciling Modern Machine Learning Practice and Classical Bias-Variance Trade-Off Abstract
10/23/23, 2:30pm MATH 501 Jacob Maibach Effective sample size: a measure of individual uncettainty in predictions Journal Paper

Spring 2023 Schedule

Date/Time Presenter Title/Topic Additional Information
02/06/2023, 2:30pm Jeffrey Mei Intro to Artificial Neural Networks Slides, Elements of Statistical Learning
02/27/2023, 2:30pm Cody Melcher Intro to PAC (Probably Approximately Correct) Learning Slides, Chapters 2/3 of "Understanding Machine Learning: From Theory to Algorithms” by Shwartz and Ben-David
03/20/2023, 2:30pm Tim Mo Dirichlet Process Mixture Model (A nonparametric Bayesian model) Slides, Sharing Clusters Among Related Groups: Hierarchical Dirichlet Processes, Markov Chain Sampling Methods for Dirichlet Process Mixture Models
04/10/2023, 2:30pm Jacob Maibach Nonparametric Bayes: Concepts and Computation Article

 

 

 
Shudong Sun Journal Seminar