Contributing machine learning methods to better understand what determines food choices in order to improve personalized recommendation systems

The stage will be covering two main fields :
1. Data collection. Currently available datasets have the big drawback of presenting a
short temporality (consumption diaries range over no more than one week period), the student will provide a system to collect data that is effective for the purpose of the project and efficient in terms of easiness of collection.
2. Identify what drives the acceptability of proposed food substitutions. The aim is to understand what makes a food substitution proposal acceptable by a person in a given context, and given his/her past history of food consumption.
The student is expected to interact a lot with the members of the ANR project as need of the various expertise will feel necessary to bear on the project.

Lieu: 
AgroParisTech
Thématiques: 
Encadrant: 
Antoine Cornuéjols
Co-Encadrant: 
C Manfredotti, F Delaere
Référent Universitaire: 
n/a
Fichier Descriptif: 
Attribué: 
No
Année: 
2 019

User login