Skip to Main Content
Practical Neural Network Recipies in C++
book

Practical Neural Network Recipies in C++

by Masters
June 2014
Intermediate to advanced content levelIntermediate to advanced
493 pages
20h 30m
English
Morgan Kaufmann
Content preview from Practical Neural Network Recipies in C++
Multilayer Feedforward Networks
111
Within that
loop,
we include
an
insurance policy against getting
stuck
at a
saddle point.
If
the directional minimization
is
not effective,
first
try
minimizing directly
in the
gradient direction.
If
that fails,
try
some random directions. Only give
up if all
fail.
Subroutine find_grad uses
the
backpropagation Equations
6-12
and
6-14 to
compute
the
gradient. Subroutine gamma computes that
constant
via the
formula
=
(c-8) 'e 6-15
8'8
where c is the negative gradient (which will be
g
on
the
next iteration),
and
g is the
current value
of the g
vector. Finally, subroutine
find_new_dir computes the new search
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Start your free trial

You might also like

Neural Networks with Keras Cookbook

Neural Networks with Keras Cookbook

V Kishore Ayyadevara
Scientific Computing with Python 3

Scientific Computing with Python 3

Claus Führer, Claus Fuhrer, Jan Erik Solem, Olivier Verdier
Mastering Java Machine Learning

Mastering Java Machine Learning

Uday Kamath, Krishna Choppella

Publisher Resources

ISBN: 9780080514338