A modeling workflow follows ordered steps: frame the problem and the metric, split the data, establish a baseline, build the features, train and compare models by cross-validation, tune the hyperparameters, then evaluate once on the test set.
A modeling workflow follows ordered steps: frame the problem and the metric, split the data, establish a baseline, build the features, train and compare models by cross-validation, tune the hyperparameters, then evaluate once on the test set.