AutoML: From Theory to Practice

Hey everyone,

We are super excited to own & restart this amazing community that Noam Maloletnik (Malali) Orion Talmi and Erez Sali built and what better way to celebrate then with a brand new meetup? Join us on May 2nd to hear from Blaz Skrlj, Phd in AutoML research and ML engineer at Outbrain followed by Noam Bresiz, CTO and Co-Founder of [Pecan.ai](https://pecan.ai/).

Agenda:

19:30 - 20:15 Evolution of AutoML: From Tables to Images, Blaž Škrlj
20:15 - 21:00 Automated Data Pipelines for ML: Challenges and how we overcome them, Noam Bresiz

Due to the covid situation, we are hosting this virtually via zoom.

# Evolution of AutoML: From Tables to Images

AutoML systems such as TPOT, auto-sklearn and similar are becoming indispensable when building modern data mining pipelines that consider different data types, from images to tabular data. This talk will present an overview of some of the main ideas emerging from the young field of design and development of AutoML systems, their impact, and future directions.

# Speaker Bio

Blaž Škrlj is a machine learning researcher active in the areas of AutoML and representation learning. During his PhD at the Jožef Stefan International Postgraduate School he worked on low-resource AutoML and its applications to natural language processing, graph-based machine learning and compressibility of latent representations. Currently, Blaž is part of Outbrain’s AutoML team, where he is exploring the limits of AutoML for large-scale recommendation.

# Automated Data Pipelines for ML: Challenges and how we overcome them

Data-centric AI, the discipline of systematically engineering the data needed to successfully build an AI system, is a leading new trend in ML research. AutoML systems should focus on the issue of data prep, not only on modeling. Noam will share the real life challenges of building automated data prep pipelines, and ways to overcome them.

# Speaker Bio

Noam is a seasoned leader in the machine learning and data analytics industry, with extensive background in software engineering. Noam is the co-founder of Pecan, a predictive analytics startup, and holds a Phd in Computational Neuroscience from Tel Aviv University.