Quiver Approaches to Machine Learning

Time:13:30 - 15:05, Tue,Thu, 9/13/2022 - 10/27/2022

Venue:Venue: 1110 Zoom: 928 682 9093 PW: BIMSA

Speaker:William Donovan (Associate Professor)

Record: Yes

Level: Graduate

Language: English


Some representation theory or algebraic geometry would be helpful.


Quiver representations are a useful tool in diverse areas of algebra and geometry. Recently, they have furthermore been used to describe and analyze neural networks. I will introduce quivers, their representations, and a range of applications, including to the theory of machine learning.



1. Quivers, quiver representations, path algebras, modules;

2. continuous groups, Grassmannians, homogeneous spaces, quiver flag varieties, moduli spaces;

3. neural networks, data flow, activation functions, gradient flow;

4. universal approximation theorems, further topics.

Lecturer Intro

Will Donovan joined Yau MSC, Tsinghua U in 2018. Since 2021 he is an Associate Professor, and Adjunct Associate Professor at BIMSA. His focus is geometry, in particular applying ideas from physics and noncommutative algebra to study varieties, using tools of homological algebra and category theory. He studied at Cambridge U, completed his PhD at Imperial College London, and was postdoctoral researcher at Edinburgh U, UK. From 2014-18 he was research fellow at Kavli IPMU, U Tokyo, where he is now Visiting Associate Scientist. His work is published in journals including Communications in Mathematical Physics and Duke Mathematical Journal. He is supported by China Thousand Talents Plan, and received a Japan Society for Promotion of Science Young Scientist grant award.

Lecturer Email:

TA: Dr. Hamed Adami,

DATESeptember 9, 2022
Related News
    • 0

      Bayesian machine learning

      Record: YesLevel: GraduateLanguage: EnglishPrerequisiteProbability theory, Mathematical statistics, Machine learningAbstractProbabilistic approach in machine and deep learning leads to principled solutions. It provides explainable decisions and new ways for improving of existing approaches. Bayesian machine learning consists of probabilistic approaches that rely on Bayes formula. It can help in...

    • 1

      The Application of Machine Learning Methods to the Solution of Partial Differential Equations II

      Record: NoLevel: UndergraduateLanguage: ChinesePrerequisiteBasic knowledge on numerical methods for partial differential equations and machine learning methodsAbstractThis course reviews the publications of the recent years on using machine learning methods to solve partial differential equations, such as Physics Informed Neural Network (PINN). The course will cover the materials on forward me...