Jelenlegi hely

Intézeti szeminárium

Félév:

2017/18 I. félév

Helyszín:

Árpád tér 2. II. em. 220. sz.

Dátum:

2017-10-10

Időpont:

14:00-15:00

Előadó:

Benczúr András (MTA SZTAKI, Budapest)

Cím:

Recommender systems by traditional and online machine learning

Absztrakt:

Recommender systems have to serve in online environments which can be

highly non-stationary. Traditional recommender algorithms may periodically

rebuild their models, but they cannot adjust to quick changes in trends

caused by timely information. In our latest experiments, we observed

that even a simple, but online trained recommender model can perform

significantly better than its batch version. In the presentation,

I will show online learning based recommender algorithms that can

efficiently handle non-stationary data sets. I will discuss evaluation

results over eight publicly available data sets. As part of our results,

I will present our open source C++ recommender system with a scikit-learn

style Python API, which is particularly suited for practical courses in

recommender systems.