Skip to main content

xapharius phoenix

xapharius phoenix's Public Library

  • More formally, do a depth first traversal of the tree and evaluate  inherited attributes on the way down and synthetic attributes on the  way up.  This corresponds to a an Euler-tour traversal

  • Neural networks of this type can have as inputs any real numbers, and they have a real number as output. For regression, it is typical for the output units to be a linear function of their inputs. For classification it is typical for the output to be a sigmoid function of its inputs (because there is no point in predicting a value outside of [0,1]). For the hidden layers, there is no point in having their output be a linear function of their inputs because a linear function of a linear function is a linear function; adding the extra layers gives no added functionality. The output of each hidden unit is thus a squashed linear function of its inputs.

  • ntul financiar in ultimele 180 zile (13.06.2013 - 10.12.2013
1 - 20 of 33 Next ›
20 items/page

Highlighter, Sticky notes, Tagging, Groups and Network: integrated suite dramatically boosting research productivity. Learn more »

Join Diigo