UPDATE - Event will be virtual and will start at 17:30. Event will be in English.

Please use this link to register:
https://imperva.zoom.us/meeting/register/tJYodO6tpzgjHNLtnN7EpXpAywyEZzt0pVER
After registration you will get the meeting link by mail.

Come to learn about 3 different data science and machine learning use case from 3 different security companies

Agenda:
17:30 - Catching IVT in ad-tech: When bots just want to watch ads - Lior Fisch
18:00 - Visual Detection of phishing websites - Shlomi Tsur
18:30 - Scaling cloud DDoS protection using AI - Johnathan Azaria

Catching IVT in ad-tech: When bots just want to watch ads

The online ad-tech industry is a surprisingly complex, multi-player ecosystem that we all interact with on a daily basis. Due to its massive annual turnover, the stakes are high and it's no surprise that ad fraud abounds, claiming billions of dollars each year.

In this talk I'll give an overview of the ad-tech ecosystem and of the most common fraud schemes that exist in it, and outline the challenges and techniques of detecting IVT (Invalid Traffic) in this context, which share many common aspects with but also bear some subtle differences from bot detection in the "normal" web security scenario. I'll also provide a peek into our work at Moat (Oracle Data Cloud) by describing some of our publicized successes in this field.

Lior Fisch, Principal Data Scientist at Oracle Data Cloud

Visual Detection of phishing websites

Visual Detection has become a common practice for detecting phishing websites. Attackers have adopted as well and are utilizing evading measures while trying to look authentic to the end user. In this talk we will review some of the evading tactics as well as ways to counter them.

Shlomi Tsur, Researcher at IRONSCALES

Scaling cloud DDoS protection using AI

A crucial step in DDoS mitigation is establishing a baseline. The standard method requires expert knowledge and is a painstaking, timely process which isn’t scalable and prone to human errors. In this talk we’ll describe the process we went through to create a new, data driven method that learns how to establish a baseline using the previous decisions made by experts. The new method is scalable, reliable and easily monitored.

Johnathan Azaria, Data Scientist at Imperva