You don't have to be excellent at maths to manage in ML but you certainly need to be comfortable with statistics and general numeric competency. You don't need to be able to invent new algorithms based on mathematical principles, but given a new ML model with e g. a description on a Medium article with the core equations and techniques shown, should be comfortable in digesting how the model works from a high level mathematical point. You need to easily understand concepts like recall, precision, F1 metrics, MSE, gradient descent, correlation, linear algebra, etc.
You cannot treat ML as a black box, it needs rigorous engineering for success. even if the final code might be no more than 20 lines of python.
You cannot treat ML as a black box, it needs rigorous engineering for success. even if the final code might be no more than 20 lines of python.