PyData TLV meetup #3

This will be our third amazing PyData event in Israel! It will include four interesting lectures by industry experts, mingling and sharing :)

Notice the event location has changed to Taboola's Offices at Habursa Ramat Gan (Taboola, Atrium Tower, floor 32 ( https://maps.google.com/maps?f=q&hl=en&q=2+Jabotinsky+Street%2C+Ramat+Gan+Parking+at+the+Maayan+parking+lot+%2825+NIS+after+5+pm%29%2C+Tel+Aviv-Yafo%2C+il ), 2 Jabotinsky Street, Ramat Gan, Parking is available at the Maayan parking lot (25 NIS after 5 pm)).

• 18:00 - 18:30 - Gathering, snacks, mingling

• 18:30 - 20:30 - Four interesting and practical data science talks

Generative Adversarial Networks, Victoria Mazo, PhD, Deep Learning Researcher at Zebra Medical

In recent years, Deep Learning has emerged as a powerful approach to learning representations directly from labelled data. In the last year, unsupervised learning has made a big step forward - Generative Adversarial Networks (GANs) have emerged as a promising framework for unsupervised learning. GANs consist of a generator and a discriminator, which learn together by pursuing competing goals. A generator function learns to synthesize samples that best resemble some dataset, while a discriminator function learns to distinguish between samples drawn from the dataset and samples synthesized by the generator. From a conceptual perspective, adversarial training is fascinating because it bypasses the need of loss functions in learning, and opens the door to new ways of regularizing (as well as fooling or attacking) learning machines. In this talk I will explain the GANs basics (and some tips for their training), show examples of their usage for image-to-image mappings and mention how GANs can be useful in the field of Medical Imaging.

Introduction to algorithmic trading, Andrew Kreimer, Founder and Quant at Algonell

More than 80% of daily traded volume is done by computers. Algorithmic Trading dominates the markets and Quantitative Finance is one of the hottest topics nowadays. This presentation is a fast and light introduction to the world of Algorithmic Trading in Python. You will get guidelines and tools to save you time and money. You will learn how to get the data, create trading models, evaluate and trade them in Python. This is your starting point to invest your money wisely with Python based technologies.

Easy Spark: Exploiting large datasets for multi-class classification, Victor Makarenkov, NLP researcher at Ben Gurion University.

Apache Spark provides an interface for programming entire clusters with implicit data parallelism and fault tolerance. At first glance, it seems that getting started with programming the Hadoop eco-system is quite cumbersome, and not so user-friendly for a data scientist or a machine learning specialist. In this talk I will briefly introduce Apache Spark, and its programming paradigm. I will show how to easily execute a distributed training of the common multi-class classifiers (naïve Bayes, random forest, logistic regression), without installing a single virtual machine, virtual box or a docker. I will share my experience of managing long-term software projects which are based on the Hadoop technology for data storage, extraction and transformation.

Working with TensorFlow - tips learned from Spark, Tal Franji, Big Data & Spark Expert

What tips can we use with TensorFlow taken from Spark/big-data experience? Examples of working with cloud machines, experimenting, iterating, interactive work, monitoring and visualisation will be shown. Some script goodies on github will also be provided for working with EC2 and Jupyter. The event is sponsored by Taboola.