Imagine you want to order food in a restaurant in a foreign country. Wouldn’t it be great if you could just ask a digital device which menu option you would enjoy the most? It’s not that straightforward - preferences are personal. If you are on a diet, you might want a low fat option. If you are vegetarian, you want to avoid meat. To help you to make a good decision, a device needs to learn all your personal preferences. This takes a lot of time and might even be impossible.
To get around this problem, Anne-Marie George is approaching the issue from another direction, known as ‘Preference Inference’. By asking questions such as ‘which one do you like best: Shepherd’s Pie, Seafood Chowder or Stew?’, it is possible to deduce a lot of information about your preferences.
This concept of Preference Inference can be applied to many decision types. Anne-Marie gets a picture of which criteria are important to you by excluding those that don’t coincide with your statement. This approach can be applied in a variety of decision-making environments
Insight Researchers: Anne-Marie GeorgeOrganising Body: Tenth International Conference on Scalable Uncertainty Management, SUM16, Event: Tenth International Conference on Scalable Uncertainty Management, SUM16, Venue: Château de Valrose, NiceDate: 21/09/2016Event Type: ConferencePresentation Type: Paper
Insight Researchers: Abdul Razak, Anne-Marie George, Nic WilsonEvent: The annual IEEE International Conference on Tools with Artificial Intelligence (ICTAI), Venue: Vietri Sul MareDate: 09/11/2015Event Type: ConferencePresentation Type: Oral Presentation
Insight Researchers: Abdul Razak, Anne-Marie George, Nic WilsonOrganising Body: 5th Workshop on Algorithmic issues for Inference in Graphical Models (AIGM), Event: 5th Workshop on Algorithmic issues for Inference in Graphical Models (AIGM), Venue: Université Paris DescartesDate: 28/09/2015Event Type: WorkshopPresentation Type: Oral Presentation