
Vladimir Vapnik: Statistical Learning
Lex Fridman Podcast
Do You Think It's Interesting Because Its Not Productive?
Understanding the nature of functions and algorithms often involves grappling with challenging principles that may not seem immediately useful. Judging something solely by extreme cases—either the worst or the best—fails to offer a comprehensive view of what lies in between, which is essential for a fuller understanding. In real-world contexts, situations often exist on a long tail, where outliers can significantly affect outcomes, complicating the task of creating a universal model. Relying solely on the law of large numbers can lead to misleading interpretations, as it doesn't account for the variance inherent in many real-world applications. Grasping these concepts provides a deeper insight into the complexities within statistical models and their application in various fields.


