Team Members:
U2110075 Rashid Zaynutdinov - Team Leader
U2110073 Omonjon Yokubjonov
U2110072 Samandar Xolmirzayev
U2110068 Ulugbek Utambetov
An Neural Network that plays Flappy Bird! Using the NEAT python module. NEAT Algorithm - Natural Selection Algorithm.
https://neat-python.readthedocs.io/en/latest/config_file.html
https://nn.cs.utexas.edu/stanley.cec02.pdf
https://medium.com/analytics-vidhya/how-i-built-an-ai-to-play-flappy-bird-81b672b66521
Run flappy_bird.py and watch an AI built - start training itself to play the game of Flappy Bird!
NEAT tries to create/delete and construct the best topology, by adjusting the weights and biases for each of the possible paths and their allocated nodes
The following Neural Network, heavily is dependent on the following Input variables:
A Bird either Jumps or it doesn't
TanH Activation Fucntion is chosen to make it easy to tell wheter a bird did or did not Jump
Progress and distance that is being covered by a bird
Set to be 30, as it is a massive treshhold, upon reaching which it is statistically impossible under the normal Flappy-Bird game mechanics - for the best-performning bird not to adapt to the gameplay and for the latter in the context of a fore-mentioned scenario to exist.
Info provided by Rashid Zaynutdinov | U2110075
For Educational Purposes Only !!!