07 06 Project 2 Multi Armed Bandits Algorithm

The Multi-Armed Bandit algorithm and its variants (Epsilon Greedy, Epsilon Greedy with Decay, Softmax Exploration) help to build live-learning intelligent agents that can take optimum actions under uncertainties. This has applications in A/B testing, online search, e-commerce, online advertisement placements, clinical trials of new drugs, etc. Use this link to access the notebook : https://colab.research.google.com/drive/1z9HL5cvA8xl-gvUW1fxKEg482r7gT2Y3?usp=sharing