Create $3$ sub$-$datasets with the top $10$ most important features given different

label sets:

$1.$NoEvents$,$Others$$

$2.$Attack$,$Others$$

$3.$Natural$,$Others$$

$***$Note: $$Others$$ will be a new label created by you, which should be assigned

to the data instances which do not belong to the underlined category. For example,

the data instances labeled as $$Others$$ in the $1$st label set are indeed the data

instances which were originally labeled as $$Attack$$ or $$Natural$.$