Codette 4 Cloud - Chapter 3

We continue the Codette 4 Cloud meetup series (previously titled Codette 4 Web) in partnership with Adobe Romania. This meetup series is dedicated to Cloud, Artificial Intelligence, Big Data and web services topics.

Please RSVP using your full name, so we can add you on the participants list that will be allowed access in Anchor Plaza.

EVENT AGENDA

18:30 - 19:00 Registration

19:00 - 19:15 Opening Notes

19:15 - 19:45 "Building Machine Learning models that could predict the next Bitcoin", Ruxandra Burtica - Software Engineer Adobe, Antreprenor

19:45 - 20:15 "Object Detection - Retele neurale pentru detectia obiectelor in imagini. Modelul Faster RCNN. Limitari ale acestui model.", Tudor Virgil - Programator UiPath

20:15 - 21:00 Networking

The presentations will be held in Romanian.

ABOUT THE SPEAKERS AND PRESENTATIONS
Ruxandra Burtica (Software Engineer @ Adobe, Entrepreneur)

Ruxandra Burtica is a software engineer, who is passionate about data engineering, architecting systems for scale, and machine learning. She spent more than 7 years working with data on multiple roles, from building data warehouse solutions at IBM, to building analytics products. She was part of the team that build the backend data pipeline for uberVU, which was later acquired by HootSuite. Right after that, she joined a U.S.-based startup, building their backend for two iterations of their analytics product. Then she co-founded an infrastructure startup where she served as the CTO, where all important decisions about product were based on user's actions. Currently she’s working at Adobe, building machine learning products.

About Ruxandra's presentation

"The Initial Coin Offering has disrupted the VC ecosystem, disinter-mediating the parties that acted as market makers and creating an ecosystem for startups to raise money directly from interested users. Only in the past few months, there have been a couple of hundreds of ICO's that happened. The key factors for determining the success rates of an ICO before the pre-ICO sales seems to be difficult to predict. I've collected data from various public sources and tried different approaches to predicting the success rate of a new ICO. I'm currently measuring the success rate of an ICO by the amount of money raised, but we will discuss alternatives.

I'll be using a Jupyter notebook as a better way to showcase the steps taken, and we'll go through all the steps needed to build a machine learning project, from data collection, analysis and cleaning to building the model, fine tuning and presenting a couple of use-cases. I'll leave you with a couple of ideas for improvements that can be made."

Tudor Virgil (Programmer @ UiPath)

struct Bio{
String Nume = "Tudor Virgil";
UInt32 Age = 28;
String Profession = "Programmer";
String currentWorkplace = "UiPath, Machine Learning Development";
String[] previousWorkplaces = new[]{"Ubisoft", "King Games"};
String[] studies = {"Faculty of Informatics at the University of Bucharest"};
String[] hobbies = {"Video games","reading"};
};