Abstract
Sandris Murins
Abstract
Presentation will share insights on pilot project of using machine learning (ML) to personalize diet plans. The pilot project is run by Europe Union based online weight management company Science22.com. It is developed by analyzing their data on 700 000 created diets plans. The goal of the project was to find ways to maximize lifetime value of users. During pilot project was found that personalization of diet plans by 20% increases the probability that users will meet their weight management expectations. Thus data scientists created model that with 72% probability predicts if a meal will be liked by a user or not. Prediction is based on the properties of the meal and social demographics data of users.