What are "weighted examples"

User137200

Hi!! When I look at the capabilities of a Machine Learning operator with "Show Operator Info" the term "weighted examples" appears in the list. For example, the Linear Regression operator says to accept weighted examples, but K-NN does not. I wish I could understand. Thanks in advance

MartinLiebig

Hi,

you can add a column of role weight with numbers of it. if you have a 2 in this, the algorithm gives it a statistical weight of 2. So its very similar to duplicating the row (not the same). This way you can force the algorithm to treat some examples as more important than others.

Best,

Martin