"Probability - Statistics Seminar in January 2024" is going to be organized with a report by Dr. Nguyen Ngoc Dung.

Dr. Nguyen Ngoc Dung is currently a postdoctoral researcher at the Department of Statistical Sciences, University of Padova, Italy. She is a former student of the Department of Mathematics - Informatics, Academic year 2013. She defended her thesis titled "Model selection for colored graphical models for paired data", in May 2022. Current research fields Dr. Dung's current work is Gaussian graphical models for paired data, high-dimensional mixture of experts.

Undergradate, Master and Ph.D. students and interested people are invited to participate the seminar.

• • Time: 2:30 p.m. to 3:30 p.m. Thursday, January 25, 2024.
  • Location: Room F207, Campus 1- University of Science(227 Nguyen Van Cu St., District 5, Ho Chi Minh City).

Summary of the report is as follows:

Gaussian Graphical Model is a statistical model that represents the conditional dependence structure between random variables using a graph. The Gaussian graph model is also known as Gaussian Markov random fields or Gaussian network. The structure of a Gauss graph basically consists of two parts: nodes and the edges, in which each node corresponds to a random variable, and each edge represents the conditional dependence between nodes (random variables). This dependence can be directed or undirected. The Gauss graph model is widely applied in various scientific fields including: biology, finance and computer science.

This talk aims to briefly introduce the Gauss graph model and related issues, from theoretical foundations to applications. The main content includes the following sections:

• • Introduction to the Gauss graph model;
  • Application of Gauss graph model (some examples in biology, finance, ...);
  • Select Gauss graph model (Lasso graph model, stepwise backward selection);
  • Introduction to the Gauss graph model with symmetry conditions;
  • Selection of Gauss graph model with symmetry condition (fused Lasso, stepwise backward selection);
  • Applying Gauss graph model with symmetry condition for paired data.

Scan the QR code below or click on the link before January 24, 2024 to register to attend.