編譯：
sourceforge上的FANN庫帶VS2010的工程，我機器上裝的VS2005，用不了，愁人，只能手動創建工程了，編譯不過，度娘不管用，翻牆麻煩，用雅虎搜到一個工程的創建配置，調整配置試一下，果然好用，在這裏備份一下：
1) 創建一個工程，可以叫"vs2005test".包含所有的庫文件(.h和.c)，除了這幾個文件(這點很重要)： floatfann.h, floatfann.c, fixedfann.h, fixedfann.c, doublefann.h , doublefann.c。
2) "工程屬性 -> C/C++ -> 常規 ->附加包含目錄"添加include路徑。
3) 添加預處理器定義: (工程屬性->C/C++ -> 預處理器 -> 預處理器定義)
_CRT_SECURE_NO_WARNINGS
_USRDLL
FANN_DLL_EXPORTS
WIN32
_WINDOWS
_DEBUG
_CRT_SECURE_NO_DEPRECATE
4) 配置 工程屬性 -> 常規 -> 配置類型 = 動態庫 (.dll)
5) 編譯.ok!
測試：
調試的話，再建個控制檯工程，包含該dll和庫，添加main.cpp，測試代碼如下：
#include <stdio.h>
#include "floatfann.h"
int main()
{
printf("Please Enter Command:(\"t\":Train,\"c\":Calculate,\"v\":Verification.)\n");
char c = getchar();
if('t' == c)
{
const unsigned int num_input = 2;
const unsigned int num_output = 1;
const unsigned int num_layers = 3;
const unsigned int num_neurons_hidden = 1024;
const float desired_error = (const float) 0.001;
const unsigned int max_epochs = 500000;
const unsigned int epochs_between_reports = 1000;
struct fann *ann = fann_create_standard(num_layers, num_input, num_neurons_hidden, num_output);
//fann_set_activation_function_hidden(ann, FANN_SIGMOID_SYMMETRIC);
//fann_set_activation_function_output(ann, FANN_SIGMOID_SYMMETRIC);
fann_train_on_file(ann, "train.data", max_epochs, epochs_between_reports, desired_error);
fann_save(ann, "test.net");
fann_destroy(ann);
}
else if('c' == c)
{
fann_type *calc_out;
fann_type input[2];
struct fann *ann = fann_create_from_file("test.net");
input[0] = 1;
input[1] = 1;
calc_out = fann_run(ann, input);
printf("xor test (%f,%f) -> %f\n", input[0], input[1], calc_out[0]);
fann_destroy(ann);
}
else if('v' == c)
{
struct fann *ann = fann_create_from_file("test.net");
struct fann_train_data *data = fann_read_train_from_file("verification.data");
fann_reset_errstr(reinterpret_cast<struct fann_error *>(ann));
for (int i=0;i != data->num_data;i++)
{
fann_test(ann,data->input[i],data->output[i]);
}
printf("Mean Square Error:%f\n",fann_get_errno(reinterpret_cast<struct fann_error *>(ann)));
fann_destroy(ann);
}
return 0;
}