Past Meetup

Current Trends in Enterprise Data Science

This Meetup is past

147 people went

Location image of event venue

Details

Hello wonderful big data developers and enthusiasts. We hope this email finds everyone well!

We are excited to announce our first event for this summer that will take place on Thursday, June 21st at 7pm. Our venue is once again the ground floor of The Cube Athens.

Our speakers will be Nikolaos Vasiloglou (https://linkedin.com/in/vasiloglou/) and Christos Malliopoulos (https://linkedin.com/in/cmalliopoulos/) from MLTrain (https://mltrain.cc/), who will be talking to us about Current Trends in Enterprise Data Science, with a focus on data privacy, representation and neural computation frameworks.

Adrianos | Euangelos | Stavros

About the Talks:
1st Talk:

Privacy, Security, and Ethics in Data Science

When a data scientist works on datasets she/he is focusing on the scientific side of the problem. In many cases though data contains private and sensitive information that the data scientist might not even be allowed to see. In the first part of this talk we will explore methods and techniques of automatic preprocessing of the data that allow the data scientist to create models without direct access to private and sensitive data. In the second part we will explore the biases that ML algorithms can have that raise ethical issues. How do you check that your classifier is not a racist?

2nd Talk:
Training Neural Networks with Enterprise Relational Data

A basic property of enterprise data is their qualitative nature. We usually handle data that represent categories rather than quantities. This is not of major concern when we employ tree-type methods for inference (forests and gradient boosted trees). Model-based methods on the other hand generalize better than trees but are suited for scalar rather than categorical data. Traditionally we overcome this limitation by converting categories to binary variables but again, this increases dramatically the input dimensionality making the learning task harder. In the talk we explain vector embeddings of categorical variables and how we can use them to train a feed-forward neural network with Tensorflow.

Speakers:

Nikolaos Vasiloglou holds a Diploma in Electrical and Computer Engineering from the National Technical University of Athens and a PhD from the department of Electrical and Computer Engineering at Georgia Institute of Technology. His thesis was focused on scalable machine learning over massive datasets.

After graduating from Georgia Tech he founded Analytics1305 LLC and Ismion Inc. He has architected and developed the PaperBoat machine learning library which has been successfully integrated and used in the LogicBlox and HPCCSystems platforms.

Currently he works as a machine learning consultant for Symantec and Infor focusing on Google's TensorFlow and has been active in developing the syllabus for a series of TensorFlow training events. His work has resulted in patents and production systems.

Christos Malliopoulos holds a Diploma in Electrical and Computer Engineering, an MSc (summa cum laude) in Probability and Statistics and a PhD (summa cum laude) in signal processing and machine learning, all from the National Technical University of Athens.

He has worked as a research scientist at the "Institute for Language and Speech Processing" of "Athena research center" and, as BI specialist and later as the manager of the BI department of Hellenic Telecommunications Organization, a subsidiary of Deutsche Telekom AG.

He has been a consulting contractor of the data science group of Logicblox Inc. Currently he works as a machine-learning consultant for Infor Inc. focusing on declarative numerical optimization frameworks and in-database machine learning.

• Sponsors
- Intracom Telecom : [http://www.intracom-telecom.com/]
- MLTrain : [https://mltrain.cc/]
- efood : [https://www.e-food.gr/]

• Schedule
7:00 - Socialize
7:25 - Welcome
7:30 - 1st Talk
8:15 - 2nd Talk
9:00 - Drinks & Pizzas

We are always looking for speakers for our meetups. If you would like to give a talk please contact.