Project 2 Genetic Programming Written by: TEAM NLP Cristopher Hernandez Suyash Singh Contact: cristopherh@csu.fullerton.edu suyash.vardhansingh@csu.fullerton.edu Class: CPSC-481-03 Introduction: The goal of this project is to build a lisp program which attempts to create arithmetic expressions that describe a 3D surface. The expressions (a population of 50) are formed randomly initially, then selectively 'bred' and mutated using in genetic programming fashion for 50 generations. Each expression is scored based on a series of given sample points which relate to the 3D surface. External Requirements: None, aside from Lisp Build: Executable from the command line without compilation. However, a compiled version can be created using: (COMPILE-FILE "project2.lisp") In a Common Lisp REPL Installation: files list: project2.lisp Other Files: average_graph.png - generated graph of average scores over 50 generations best_graph.png - generated graph of best scores over 50 generations wors_graph.png - generated graph of worst scores over 50 generations testoutput.txt - raw output from generated data test_data.xlsx - excel formatted data Setup: This solution was written and tested with GNU Common Lisp for Windows. CLISP for windows can be downloaded and installed here: https://sourceforge.net/projects/clisp/files/ Usage: run on command line: clisp project2.lisp load compiled .fas in REPL: (LOAD "project2.fas") Extra Features: None Bugs: No known bugs at this time.
CristopherH95/CPSC-481-Project-2
Genetic Programming Project for CPSC-481 Artificial Intelligence
Common Lisp