Tensorflow and Sparklyr: Scaling Deep Learning and R to the Big Data ecosystem
Details
https://secure.meetupstatic.com/photos/event/d/f/0/7/600_460017095.jpeg
Please pay attention: the presented slides will be in english but the talk will be in Italian. An offline recording in English will be made available after the event on our YouTube channel.
Agenda:
18:30: Doors opening
19:00 Tech talks
Image Caption Generation: Intro to Distributed Tensorflow and Distributed Scoring with APACHE Spark
Abstract:
Deep Learning and its main development library, TENSORFLOW, are changing the way Data Science is perceived: being able to master those concepts and technologies is keep gaining more importance.This talk will introduce the modeling and technological issues that prevent scalability with Deep Learning applications, and describes approaches that can be used to scale-out in a distributed fashion.For this purpose a deep dive on the Image Caption Generation model provided by Tensorflow will be presented, applying it to a large set of data.
Bio:
Since November 2015, Luca Grazioli is a Data Scientist at ICTeam. His passion for Machine Learning arose already during his university studies, after which he could finalize his research path with the definition of a Knowledge Engineering model and the publication of the related paper “Modeling and Understanding time-evolving Scenarios” (http://www.iiisci.org/journal/CV$/sci/pdfs/SA268SN15.pdf )
His main focus in ICTeam is to address Data Science issues in Big Data environments, reason that prompted him to specialize in Big Data technologies applied in Data Analysis and Machine Learning contexts.
Recently, he has dedicated himself to the world of Deep Learning: how to approach it and how it integrate it with the Big Data ecosystems.
Sparklyr: Big Data enabler for R users
Abstract:
This talk will introduce the new package sparklyr, which takes R users to a new level. In fact, it allows performing big data analysis in Spark through R, taking advantage of the useful package dplyr.
Then, an application on the famous 1 billion NYC taxi data as described by Todd Schneider will be showed, performing some data manipulation and running machine-learning algorithms at scale.
Bio:
Serena Signorelli is a PhD student in Analytics for Economics and Business at the University of Bergamo. Her main research focus is about the use of big data sources in official statistics. During her PhD, she spent six months as a trainee at the Big Data task force of Eurostat, the statistical office of the European Union, where she fell in love with data science.
She started working as a data scientist at ICTeam in February 2017.
---
Remember also to buy your free ticket on the Eventbrite event page:
Just the RSVP at this meetup page will not guarantee your seat.
---
If you are interested in doing a lightning talk or want to propose an event, hackathon, panel discussion, roundtable or any sort of initiative please submit your proposal at https://goo.gl/forms/mprwZfujXSAw9emk1 (https://goo.gl/forms/mprwZfujXSAw9emk1) or get in touch with one of the member of the staff.
---
About the host:
ICTeam, with head offices in Bergamo (Grassobbio), operates in the field of ICT (Information and Communication Technology). Founded in 1999 and with more than 100 employees, the company’s core business has focused on offering system integration services, application solutions and cloud services, supporting customers throughout all the various software solution development and management phases.
The Software Development and System Integration division offer services and solutions with high added value. The division focuses mainly on the design and integration of systems, regardless of their application context. Activities concentrate mainly on 3 different operating areas:
- System Integration- Performance Assessment & Optimisation- Data Analysis and Big Data
ICTeam has gained in the last years strong experience in the Big Data context becoming a Certified Silver Partner Cloudera and a Certified Dataiku Data Science Studio Partner
More informations are available on our site http://www.icteam.it
Tensorflow and Sparklyr: Scaling Deep Learning and R to the Big Data ecosystem