home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Club Amiga de Montreal - CAM
/
CAM_CD_1.iso
/
files
/
618a.lha
/
NeuralNetwork
/
C_programmers.readme
next >
Wrap
Text File
|
1992-03-08
|
2KB
|
47 lines
This document is for C programmers who are not familiar with the
C++ constructs. The C code uses a 'class' design to implement the
Neural network.
A class is an object which has data and functions which operate on the
data. The data is generally private so users of a class have no direct access
to the data except through the class functions. In this case there is a
struct Neural_net which contains all the information needed to train and test
a neural network. However, you should never directly access this data
structure. Only use the functions provided.
A class must first be constructed (initialized) which is done by calling the
class's constructor. There are several constructors for the Neural network.
Neural_net_constr (...) --> All size and learning parameters must
be specified
Neural_net_read_constr (...) --> Reads size and weights from a file and
learning parameters must be specified.
Neural_net_default_constr (...) --> Only size needs to be specified.
Learning parameters are set to default
values.
Neural_net_default_read_constr (...) --> Read size and weights from file
and learning parameters are set to
default values.
Each constructor will return a pointer to a Neural_net if successful,
otherwise it returns NULL. Now that you have a pointer to a valid Neural_net,
you may call any of the functions which operate on a Neural_net and pass as
its first parameter the pointer to the Neural_net.
When you are finished with the Neural_net, you must destroy (free) it.
To do this you just call the Neural_net destructor
Neural_net_destr (Neural_net *);
This function will free all memory associated with the Neural_net and the
pointer itself.
See the header file Neural_net.h for an explanation of the functions and
their prototypes. See the document Neural_network.doc for a full explanation
of each function and look at the programs xor_c_dbd.c and xor_c_bp.c to see
how a Neural_net is constructed, used, and destructed in a typical example.