Hello! February 27, we invite practitioners in Data Science to a mitap dedicated to user personalization. You will hear reports from representatives of the University of Singapore Management and Avito. Register for a meeting and invite colleagues. Under the cut - theses of speeches, links to the registration and video broadcast mitap.
We often encounter a huge amount of information, interacting with search engines, e-commerce systems, social networks and so on. Considering that only a part of this information will be of interest to a specific user, it is necessary to personalize interaction with such systems, recommending to users in advance information that potentially matches their preferences.
The key to good recommendations is modeling huge amounts of behavioral data that arise as a result of user interaction with online systems. The problem here is that these interactions are multimodal, that is, composed of various types of data, such as user ratings, reviews, photos, or their social interactions. In this report we will talk about data mining and machine learning methods for modeling user preferences in multimodal data and their use in creating a complete recommender system.
This report will be in English.
There is a need for a person who wants to know how to use. in. It follows that this is the case. In this talk, we’re looking at building end-to-end recommendation systems.
Vasily Lexin will tell you how to unsupervised clustering of short-lived ads using the example of Avito in order to obtain stable clusters of goods for subsequent use in recommendations algorithms. Will present an optimized EM-algorithm that can effectively handle huge amounts of data. He will tell you about methods for assessing the quality of clustering and about the applied applications of the algorithm.
18:00 - 18:30 - Registration
18:30 - 19:30 - Learning User Preferences from Multi-Modal Data - Hady W. Lauw, Maxim Tkachenko (Singapore Management University)
19:30 - 19:45 - Coffee break
19:45 - 20:30 - Clustering of volatile ads using the EM algorithm - Vasily Leksin (Avito)
Participation in the event is free, registration is required .
Address: Avito office, Lesnaya 7 .
The broadcast will be organized on the day of the event on our youtube-channel AvitoTech . Subscribe to not miss anything!
See you!
Source: https://habr.com/ru/post/349114/