Testing Your AISee For Yourself

Cross-validation is most often used to choose something, like the degree of the curve. For each degree below, the model is trained once per fold, and the purple curves are the k models. The chart collects the cross-validation error of every degree you try. Complete the three missions.

Degree of the curve: ${degree}

Number of folds: k = ${k}

The cross-validation score of a degree is the averagelargestsmallest of its fold scores, and degree 2 has the lowest one — the same choice the single validation set made in Train, Test and Validation, but now every farm had its say.