Multi Armed Bandit Algorithm W\ Deep Learning

Build using python 3.6.6 with tox 3.4.0 and tensorflow 1.12.2

Policies implemented:

Epsilon Decreasing
Epsilon First
Epsilon Greedy
EXP3
GreedyMix
Softmax
Softmax Decreasing
SoftMix
UCB1
UCB1-Tuned
UCB2

Arms:

Bernoulli
Normal