Meshless Methods in Machine Learning
When
12:30 – 1:30 p.m., Feb. 1, 2024
Where
Speaker: Carlos Brito-Loeza, Department of Mathematics, Autonomous University of Yucatán
Title: Meshless Methods in Machine Learning
Abstract: Kernels are versatile tools used in various disciplines of numerical analysis, such as approximation techniques, interpolation, and meshless methods used to solve partial differential equations. This presentation demonstrates the utilization of meshless approaches in machine learning models and the challenges that arise from the data distribution which can lead to the formation of large and dense ill-conditioned systems of equations. A variety of techniques will be used to address this problem.