Maksym Zubkov Ph.D.
Menu

Seminars

I organize collaborative reading groups and learning seminars around geometry, tensors, and neural networks.

  1. Current seminar · Summer 2026

    Online Reading Course on Tropical Geometry

    Ongoing

    About the seminar

    An international online reading course centered on Michael Joswig’s Essentials of Tropical Combinatorics. We meet to read the book collaboratively and discuss how tropical methods connect with optimization and the geometry of neural-network function spaces.

    When
    Selected Tuesdays, 2:00 PM–3:00 PM Pacific Time
    Your local time
    Detected automatically when the page loads.
    Rhythm
    Approximately every two weeks
    Format
    Online reading course
    Place
    Online
    Organizer
    Maksym Zubkov

    If you want to join, please email me at mzubkov@dgist.ac.kr.

    Published schedule

    1. Reading

      Opening meeting: Essentials of Tropical Combinatorics

      Led by Richard Lim

      Introduction to the course and the opening discussion of Michael Joswig’s Essentials of Tropical Combinatorics.

    2. Reading

      Chapter 1 of Essentials of Tropical Combinatorics

      Led by Richard Lim

      An introductory reading and discussion of Chapter 1 of Essentials of Tropical Combinatorics by Michael Joswig.

    3. Reading

      Chapter 1 of Essentials of Tropical Combinatorics — continued

      Led by Richard Lim

      Continuation of Chapter 1 of Essentials of Tropical Combinatorics by Michael Joswig.

    4. Reading

      Chapter 1 of Essentials of Tropical Combinatorics — conclusion

      Led by Anja Meyer

      Conclusion of the shared reading and discussion of Chapter 1.

Archive

Past seminars

Brief summaries of earlier reading groups and seminars.

  1. Spring2026

    Reading Group on Algorithms in Invariant Theory

    About the seminar

    This reading group focused on Section 4 of Bernd Sturmfels’ Algorithms in Invariant Theory, which develops methods for computing invariants of arbitrary polynomial representations of the general linear group GLₙ(ℂ). The goal was to understand how invariant theory can be made explicit and computable.

    Necessary background from earlier sections—including polynomial ideals, Gröbner bases, and basic invariant theory—was reviewed selectively as needed.

  2. Fall2025

    Informal Reading Group on Classical Algebraic Geometry

    About the seminar

    This reading group studied the first three chapters of Igor V. Dolgachev’s Classical Algebraic Geometry: A Modern View.

    The discussions explored connections between algebraic neural networks and tensor decompositions, including the Waring problem, Chow varieties, and simultaneous tensor decompositions. We developed a deeper understanding of the geometry of binary forms, conics, and plane cubics, especially from the perspective of apolarity and dual varieties.

  3. Fall2025

    Learning Seminar on Tensors and Causal Inference

    About the seminar

    This seminar explored the interplay between tensors, causal inference, and modern statistical methods. We studied how tools from multilinear algebra, combinatorics, and algebraic geometry reveal hidden structure in data and enable advances in inference. Topics included tensor decompositions, graphical models, and approaches to understanding complex dependencies in data, with applications ranging from causal discovery to high-dimensional learning.

  4. Summer2025

    Reading Group on Tensors and Causal Inference

    About the seminar

    This summer reading group brought together paper discussions and visitor talks on tensors, neural networks, and causal inference. Topics included cluster patterns in tensor data, identifiability of polynomial neural networks, Bayesian networks, causal discovery, tensor invariants, and characteristic imsets.

  5. Spring2024

    Math and ML Reading Group

    About the seminar

    This participant-led reading group brought together introductory and applied talks at the meeting point of mathematics and machine learning. Speakers explored polynomial neural networks, universal approximation and backpropagation, physics-informed learning, classification, sparse identification of nonlinear dynamics, transformers, comparisons between natural and artificial intelligence, neural-network methods for macroeconomic models, variational autoencoders, and large-language-model safety.