Click here to see it in action! This is a javascript applet works on genetic evolution algorithms.

• Rockets are generated with initially random paths
• Based on distance from the target, crashing, goal achieving, and time taken, each rocket gets a fitness score.
• After each lifespan (250 cycles), the old popuation dies and a new population is created from the gennome of the old population.
• Rockets with higher fitness have higher probablilty of reproducing.

• Write better fitness function to evolver faster path
• Make better information representation
• Add intelligence (neural nets?) to individual rockets so as to ensure individual learning as well as population evolution.

• Siraj Raval
• Dan (The Coding Train)

MIT