lunduniversity.lu.se

Digit@LTH

Faculty of Engineering, LTH

Denna sida på svenska This page in English

Digit@LTH: Events

CS MSc Zoom Presentation 8 June 2020!

Föredrag

From: 2020-06-08 09:00 to: 10:00
Place: https://lu-se.zoom.us/j/61795474317
Contact: birger [dot] swahn [at] cs [dot] lth [dot] se
Save event to your calendar


One Computer Science MSc thesis to be presented on 8 June 2020 via Zoom

Monday, 8 June there will be a master thesis presentation in Computer Science at Lund University, Faculty of Engineering.

The presentation will take place via Zoom at: https://lu-se.zoom.us/j/61795474317

Note to potential opponents: Register as an opponent to the presentation of your choice by sending an email to the examiner for that presentation (firstname [dot] lastname [at] cs [dot] lth [dot] se). Do not forget to specify the presentation you register for! Note that the number of opponents may be limited (often to two), so you might be forced to choose another presentation if you register too late. Registrations are individual, just as the oppositions are! More instructions are found on this page.


09.00-10.00

Presenter: Gustav Hertz
Title: Data Optimization for a Deep Learning Recommender System
Examiner: Flavius Gruian
Supervisors: Patrik Persson (LTH), Emil Joergensen (IKEA)

This thesis investigates the performance of a deep learning recommender system based on the given training data. The recommender system used in this thesis a Long-Short Term memory. First, the performance of this recommender system is defined as a combination of Precision, Catalog Coverage, and Novelty. Then the performance as a function of the available amount of training data is investigated. We conclude that depending on how one values the given performance metrics, there is an optimal amount of training data. Secondly, how to handle an excess of training data is investigated, concluding that more recent data leads to better recommender system performance. Finally, how to use data from secondary markets when there is a lack of data for training the recommender system is investigated. A similarity metric for purchase data between markets is proposed and results are promising, although more research is needed on this topic.