This is a collection of bio-inspired AI projects I took as an undergrad. They are mainly intended as a backup, but anyone in need is welcome to benefit from them. Specific descriptions follow in turn.

Genetic Algorithm for Task Scheduling Problem

This is a c++ genetic algorithm aimed at solving the task scheduling problem.

Compilation

Make targets include:

compile

run: runs single instance. Running in non-batch mode gives additional info about the best subject found.

batch: runs in batch mode. Don't forget to activate batch mode in the include/parameters.h file and to check lib/batch.sh file for preferences. Returns only the best solution score.

clean

all

Attempt at a Genetic Algorithm Framework in Ruby

Supports different reinsertion, selection and crossover methods, as well as loading individual instances from text files and a parameter configuration through the app/default_parameters.rb file. Example implemented problem is crypto-arithmetic.

reinsertion: best, elitism

selection: tourney, roulette

crossover: cyclic, pmx

I'm not a .NET guy, but I contributed heavily in this project for an Artificial Intelligence course taken together with the Master's program as a grad student. These involve some pretty advanced algorithms in biologically-inspired artificial intelligence.

Multi-objective Genetic Algorithm

This project includes implementations for mono-objective crypto-arithmetic and multi-objective task scheduling using NSGA2 and SPEA2 approaches.

Ant Colony and Particle Swarm Optimization for Traveling Salesman Problem

This project compares both methods in solving the TSP.