Why is x an array in the softmax function?
Sorry if this sounds dumb, I'm new to ML. So basically, if the softmax function is "e ^ (x - max(x)) / sum(e^(x - max(x))", I understand what e is, but then what's x? I keep seeing people make x an array, but shouldn't it just be a single number or float? Isn't that what the output would be of the summation formula? Thanks.
1 comments
This is better answered in stack overflow but x is an array. I think a better written string is e ^ (x_i - max(x)) / sum_i(e^(x_i - max(x))"
x_i is a specific element in the array x. Basically an array of real values is converted to an array of 0 to 1, with one corresponding to the largest element in x