C++ Neural Networks and Fuzzy Logic by Valluru B. Rao M&T Books, IDG Books Worldwide, Inc. ISBN: 1558515526   Pub Date: 06/01/95

Appendix A
Compiling Your Programs

All of the programs included in the book have been compiled and tested on Turbo C++, Borland C++, and Microsoft C++/Visual C++ with either the small or medium memory model. You should not have any problems in using other compilers, since standard I/O routines are used. Your target should be a DOS executable. With the backpropagation simulator of Chapters 7, 13, and 14 you may run into a memory shortage situation. You should unload any TSR (Terminate and Stay Resident) programs and/or choose smaller architectures for your networks. By going to more hidden layers with fewer neurons per layer, you may be able to reduce the overall memory requirements.

The programs in this book make heavy use of floating point calculations, and you should compile your programs to take advantage of a math coprocessor, if you have one installed in your computer.

The organization of files on the accompanying diskette are according to chapter number. You will find relevant versions of files in the corresponding chapter directory.

Most of the files are self-contained, or include other needed files in them. You will not require a makefile to build the programs. Load the main file for example for backpropagation, the backprop.cpp file, into the development environment editor for your compiler and build a .exe file. That’s it!