2.2. Data mining approach and evaluation
We will adopt a regression approach, which preserves the order of
the preferences. For instance, if the true grade is 3, then a model that
predicts 4 is better than one that predicts 7. A regression dataset D is
made up of k {1,…, N} examples, each mapping an input vector with I
input variables (x1 k ,…, xI
k) to a given target yk. The regression
performance is commonly measured by an error metric, such as the
mean absolute deviation (MAD) [