Choosing the Right Machine Learning Model

Question

When faced with a problem, how do you determine which machine learning model to choose? Do you rely on intuition, model performance metrics, or trial and error? Share your strategies for selecting the best model.

Answer 1

You could have some rough estimate based on general performance. For example, if you're dealing with sequential data and have the computation power transformers are generally state of the art so you can start from there.
Or, for general tasks and you want to easily avoid over fitting, random Forest can perform well.

Otherwise, it is often best to just try many different models as a baseline and compare their performance.