2D simulation of evolution using neural networks. REFACTOR IN PROGRESS!
Simulation creates 10 animals. Each brain of animal has four inputs( is food: up,down,left,right). 2 hidden layers. 2 output neurons telling how fast to move vertically and horizontally.
Scoring of fitness: More food animal eats higher the score, plus how much time it spent close to food.
DNA: It is their brains weights, vector of floats, the size.
Evolution: After scoring, 2 best brains continue to next generation, 7 are created by crossover with the best. Plus one is always reinitialized.
Requirements: -SFML graphic library -C++11