[Seminar]MLDS Seminar 2023-3 by Mr. Guillaume Houry (Universite Paris-Saclay), Mr. Yuxan Wan (Michigan State Univ. ), Seminar Room L5DE23

Speaker 1: Mr. Guillaume Houry, Ph. D. Student, Universite Paris-Saclay

Title: Average complexity of Persitence Homology computing in Topological Data Analysis

Speaker 2: Yuxuan Wan, Ph. D. Student, Michigan State University

Title: Defense Against Gradient Leakage Attacks via Learning to Obscure Data

[Seminar]MLDS Seminar 2023-2 by Mr. Haoyu Han (Michigan State Univ.), Mr. Weijie Liu (Zhejiang Univ.), Seminar Room L5DE23

Speaker 1: Mr. Haoyu Han, Ph. D. Student, Michigan State University

Title: Alternately Optimized Graph Neural Networks

Speaker 2: Mr. Weijie Liu, Ph. D. Student, Zhejiang University

Title: Robust Graph Dictionary Learning

[Seminar] Metric Recovery from Unweighted k-NN Graphs by Ryoma Sato (Kyoto University), Seminar Room L5D23

Metric Recovery from Unweighted k-NN Graphs by Ryoma Sato (Kyoto University)

[Seminar] MLDS Seminar 2023-1by Mr. Pengfei He(Michigan State Univ.) and Mr. Yuki Takezawa(Kyoto Univ.), Seminar Room L5DE23

Speaker 1:Mr. Pengfei HE, Ph. D. Student, Michigan State University

 Title: Probabilistic Categorical Adversarial Attack

Speaker 2:Mr. Yuki TAKEZAWA, Ph.D. Student, Kyoto University

 Title: Beyond Exponential Graph: Communication-Efficient Topologies for Decentralized Learning via Finite-time Convergence 

[Seminar] Proper Losses, Moduli of Convexity, and Surrogate Regret Bounds, by Dr. Han Bao, Kyoto University

Assistant Professor Han Bao, Kyoto University, "Proper Losses, Moduli of Convexity, and Surrogate Regret Bounds" 

[Seminar] Expected Expressivity and Gradients of Maxout Networks by Ms. Hanna Tseran, MPI

Ms. Hanna Tseran, MPI, "Expected Expressivity and Gradients of Maxout Networks" 

Lecture series "The quest for mathematical understanding of artificial intelligence", Professor Sanjeev Arora (Princeton University)

Online Lecture series "The quest for mathematical understanding of artificial intelligence" by Professor Sanjeev Arora, Computer Science at Princeton University.