(us) data drift

This article on "data drift" (by Augustin Marty) reminds us on how important the #MLCanvas method is to consider and implement the regular performance checks and updating of your deployed models.

#MLCanvas is the method, created by Louis Dorard, that enables teams, data science and business profiles, to collaborate better, around a single page. This method unite minds, at a glance, on all the ins and outs of an #ArtificialIntelligence system.

The article could also point out, as part of the ever evolutive data context, the economic equation will also have to change and be reconsidered, justifying a continuous monitoring of the deployed models. Indeed your competitors will soon realize the interest of the #AI in your market.

https://www.usinenouvelle.com/blogs/augustin-marty/le-data-drift-ou-le-cauchemar-de-l-intelligence-artificielle-en-production.N966826