Suppose I have target vector y= (y_1, y_2, ..., y_n)
where y_i in [0, inf)
for all i
e.g.
# here n = 6 and n is constant for all data
y = [
0,
0,
2,
1,
0,
1
]
In general y_i
both integers and in practice almost always are in [0,10]
. In theory, however they can take any non negative real value.
What is the best way to handle this for training and which loss should I use?
Should I treat it like classification and refractor y
so the above example becomes
# assuming zero indexed
[
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0], # 1
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0], # 1
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0
[0, 1, 0, 0, 0, 0, 0, 0, 0, 0], # 1
]