There is a commonly used convention. The subscript is reserved to indicate the elements of array and superscript to represent the iteration (over time) in neural network area.
Superscript and subscript in ANN formula
1 次查看(过去 30 天)
显示 更早的评论
Why supercript is used in the ANN training algorithms formulas instead of subscript when denoting the number of iterations? Usually in mathematics, superscripts are used to denote the exponent or power but why is it used to denote the number of iterations (isn't supposed to be subscript)?
Eg. Formula of Levenberg-Marquardt algorithm is shown in the picture attached.