The No Regrets Waiting Model: A Multi-Armed Bandit Approach to Maximizing Tips

Carol Louise Feurtalini^1,2

¹ Head Waitress at Mama Lou’s Seafood Shack

² Department of Statistical Waitressing Studies, Cranberry-Lemon University, Pittsburgh, PA, USA

Abstract

When I started waiting at Mama Lou’s Seafood Shack, home of the world famous Mama Lou’s Seafood platter for only $29.99, I knew I hit a gold mine. With that delicious food, which isn’t too filling, loads of additional apps and sea themed cocktails, the tips working here have been HUGE. I wasn’t just gonna put in as many hours as I could, I had to keep the wait staff to a MINIMUM so I can get as many tips as I can. Ole Carol here’s gonna retire early Kaching. Problem is, If I don’t pay enough attention to certain customers over another, I can’t serve all of them well enough to get those tips, let alone serve everyone enough food to rack up a huge bill and that sweet sweet %20 gratuity for larger parties while keeping Mama Lou off my back. One day it all hit me, you can formulate this as a Multi-Armed Bandit problem in which each customer is represented as a bandit arm and bada-bing-bada-boom, all I needed was a Reinforcement Learning algorithm to solve it. Using an Epsilon-Greedy, an Upper Confidence Bound, and a Thompson Sampling Algorithm, this paper will create a No-Regrets method to serve my customers for maximum tip extraction. All algorithms worked exceptionally well except for the epsilon-greedy method which lost me my job but won me the love of my life, Tony. So I’d call that a No-Regrets model.

Keywords:   Reinforcement Learning, No-Regrets, Seafood Platter, Multi-Armed Bandit Problem, Waitressing, Epsilon-Greedy, Endless Shrimp, Upper Confidence Bound, Dessert Push Estimation, Round Robin Scheduling, Thompson Sampling, Customer Prior Estimation, Alcohol Purchase Probability Chain

If you would like to read the rest of this article, it is a chapter of our new book How to Prove Anything: 30 absurd research papers no one else was brave enough to publish which can be purchased here on amazon of from Packt directly

If you enjoyed this well peer-reviewed article optimizing the act of waitressing in an easy-to-follow method I’m sure every single waiter and waitress will start using with no problems whatsoever, please like, share, and subscribe with your email, our twitter handle (@JABDE6), our facebook group here, or the Journal of Immaterial Science Subreddit for weekly content.

If you REEEEALY love the content, for the equivalent price of a Chipotle Burrito, chips and Queso, you could buy our new book Et Al with over 20 hand picked Jabde articles for your reading pleasure, it’s the perfect Christmas/Birthday gift for confusing your unsuspecting family members! Order on amazon here: https://packt.link/at4bw Please rate and review so that you can brag to your friends about having opinions or showcase your excellent taste in reading material!