Explore the intricacies of handling correlated variables in this 44-minute talk by Åsmund Rinnan. Delve into the concept of covariance, its implications, and effective strategies for managing correlated data. Learn about various techniques including orthogonalization, orthogonal signal correction, and direct orthogonalization. Examine real-world applications through example papers, with a focus on total fatty acids. Discover the importance of predictive components and data summarization in dealing with covariance. Gain valuable insights into this crucial aspect of chemometrics and machine learning, despite some audio quality issues.