DataTalks #35: Similarity Measures in ML 🧑‍🤝‍🧑📏

Our 35th DataTalks meetup will be hosted by Darrow.AI at their offices, and will focus on applications and implications of similalrity measures in machine learning!

Location: 21sth floor, Yitzhak Sade 8, Tel Aviv
🚘Note: No parking is available! The closest parking lot is @ Nitsba Tower, HaMasger St. 39.

𝗔𝗴𝗲𝗻𝗱𝗮:
🍕 18:00 - 18:20 - Mingling, etc.
🔶 18:25 - 19:10 – Distribution metrics: From GANs to data analysis
🔷 19:15 - 20:00 – Explainable similarity for Recommender Systems

***

Talks #1: Distribution metrics: From GANs to data analysis

Speaker: Uri Itai, Ph.D. Senior Data Scientist @ TRST AI

Abstract: Similarity measures for distributions attracted the attention of both theoreticians and practitioners. In this talk, we will discuss the importance of these metrics in the GAN scheme. Following this, we demonstrate the use of such metrics in other uses of machine learning such as EDA and measures of goodness of fit.

We assume basic machine learning knowledge. However, this talk will provide tools for deep understanding of the field.

***

Talks #2: Explainable similarity for Recommender Systems

Speaker: Benjamin kempinski, NLP Expert @ argmax.ml

Abstract: Recommendation systems are everywhere, with one goal in mind - improving user engagement. Studies have shown that a machine recommendation is more effective, if an explanation is given next to it. In this talk, we would cover how to learn an explainable similarity measure between items using compositional methods and we would demonstrate several use cases.