Watch a detailed lecture exploring the concept of "taskification" in data science and its relationship to creating valid benchmarks for evaluating machine learning models and systems. Dive into the methodological considerations and challenges involved in breaking down complex problems into discrete, measurable tasks while ensuring benchmark validity. Learn how proper taskification enables more accurate assessment of model performance and facilitates meaningful comparisons across different approaches.
Overview
Syllabus
Lecture starts
Taught by
UofU Data Science