Dual-SPLS - A Versatile Approach Improving PLS with Lasso Shrinkage

Explore a 42-minute webinar on dual-sPLS, a novel algorithm that enhances Partial Least Squares (PLS) with Lasso shrinkage. Presented by Louna Alsouki from the University of Claude-Bernard and University of Saint Joseph de Beyrouth, learn how this versatile approach tackles high-dimensional problems in analytical chemistry. Discover the integration of projection methods and variable selection algorithms to achieve more accurate and interpretable results. Understand the mechanics of sparse PLS and how dual-SPLS improves upon it, offering a sparser data representation while maintaining predictive accuracy. Delve into the concept of dual norms and their application in the algorithm, including the handling of multiple predictor sets related to a single response variable. Gain insights into how adaptive penalization and various norm types contribute to solving complex analytical chemistry challenges. Access accompanying slides for a comprehensive understanding of this innovative technique.