Understanding the notions of perfect classifier and random classifier
When reading performance metrics of Machine Learning algorithms, it is important to know what is the minimum and the maximum possible values of these metrics. The notions of perfect and random classifiers can give you a stronger understanding of classification performances. — This article was co-authored by Clément Côme and Charles Tremblay. It was originally published in french on our blog. In this article, we introduce three key concepts for understanding classification metrics: standard, perfect and random models.