Everything is a game: use cases for Reinforcement Learning

Reinforcement Learning (RL) is typically presented using examples of playing different games like Atari classics or physics-oriented simulators. Often one can hear opinions that RL usage in other industries and areas is very limited or challenging to implement. Such a statement is far from true - RL can be employed for similar problems as classic optimization tools, utilized by operations research (OR) for many years: production planning, scheduling, and resource optimization. With just a few mental tricks, an analyst can turn most optimization problems into a form that can be solved using these algorithms. During a presentation, a couple of such examples will be shared, as well as tips&tricks for designing your cases.