Simulates an artificial neural network. Also provides a testing program, where you can test old data or get the results for new data. Finally, supplies examples for an AND Gate and an image recognition of the digits 1 - 9. All done using FORTRAN 90.
ANN is a program which simulates an artificial neural network.
TANN is a program which tests an artificial neural network.
AGE is a program which generates training data for an AND gate.
MNISTF generates training data for image recognition of the numbers 1 - 9.
AGE and MNISTF can be found in the Examples directory, under AND Gate and Image Recognition, respectively.
To compile the source code:
-
To make an ANN executable: Type “make ANN” in the terminal.
-
To make a TANN executable: Type “make TANN” in the terminal.
-
To make an AGE executable: Type “make AGE” in the terminal.
-
To make an MNISTF executable: Type “make MNISTF” in the terminal.
-
To clean up the files afterwards: Type “make clean” in the terminal.
-
To run ANN: Type “./ANN” after making an executable, then follow the on-screen instructions (surround file names like so when instructed: “‘...’”.
-
To run TANN: Type “./TANN” after making an executable, then follow the on-screen instructions (surround file names like so when instructed: “‘...’”.
-
To run AGE: Type “./AGE” after making an executable, then follow the on-screen instructions (surround file names like so when instructed: “‘...’”.
-
To run MNISTF: Type “./MNISTF” after making an executable, then follow the on-screen instructions (surround file names like so when instructed: “‘...’”.
The training data format can be found in the file ‘Training_Data_Format.txt’.