• Deep Learning en R con MXNET

    Digital Origin

    Nota: La charla será en castellano. En este evento, haremos una introducción a las redes neuronales con una doble perspectiva teórica y práctica. Veremos una introducción a deep learning, y a las arquitecturas más utilizadas en la actualidad (CNN y RNN). R posee ya varias librerías para implementar redes neuronales, pero la librería MXNET es de la pocas librerías en R que permite ajustar modelos con redes neuronales del estado del arte actual en deep learning. Veremos también una introducción a MXNET y como implementar estos modelos con ejemplos prácticos. Presentado por Roger Borras, Estadístico en Hospital Clínic de Barcelona y miembro fundador de la asociación sin ánimo de lucro Barcelona Data Science Society. Al final de la charla abrá algo para picotear ofrecido por Digital Origin (https://www.digitalorigin.com/) que nos dará la posibilidad de conocernos mejor. Los próximos eventos relacionados a R se publicarán en el grupo "Barcelona Data Science and Machine Learning Meetup" (https://www.meetup.com/barcelona-data-science-machine-learning/). Si quieres seguir nuestros próximos eventos inscríbete allí!

  • Automatic Tools For Improving Packages

    King Offices

    For this event we'll have Maëlle Salmon ( http://www.masalmon.eu/ )as the main speaker. Maëlle is very active in the Data Science community, she writes often about R, statistics and data analysis in her blog; she is the author of several R packages and one of the organizers of the R-Ladies meetups. In this event Maëlle will explain how to improve your own R packages using "pkgdown", "goodpractice" and other tools. If you have no experience in making your own package, in the first part of the meetup you will have the chance to learn the very very basics on making an R package. This will be an hands-on event, so please bring your laptop with you or pair with someone at the event. Before the meetup is highly recommended that you: - get the package template available at https://github.com/maelle/fakepackage - Install RTools (https://cran.rstudio.com/bin/windows/Rtools/) - Install devtools, lintr from CRAN install.packages("devtools") install.packages("lintr") - Install goodpractice and pkgdown from Github: source(" https://install-github.me/MangoTheCat/goodpractice ") devtools::install_github("hadley/pkgdown") As for King's policy you will have to register at the entrance before attending the MeetUp. To speed up the registration process, please fill the form at: https://goo.gl/forms/l50xkGdRuRV9Hepn1 Thanks a lot for your help and see you at the event!

  • H2O Deep Water - Making Deep Learning Accessible to Everyone

    Universitat Pompeu Fabra, sala 52.023

    Deep Water is H2O's integration with multiple open source deep learning libraries such as TensorFlow, MXNet and Caffe. On top of the performance gains from GPU backends, Deep Water naturally inherits all H2O properties in scalability. ease of use and deployment. In this talk, I will first give an introduction to our open source H2O machine learning platform and then go through the motivation and benefits of Deep Water. After that, I will demonstrate how to build and deploy deep learning models with or without programming experience using H2O's R/Python/Flow (Web) interfaces. Bio: Jo-fai (or Joe) is a data scientist at H2O.ai. Before joining H2O, he was in the business intelligence team at Virgin Media in UK where he developed data products to enable quick and smart business decisions. He also worked remotely for Domino Data Lab in US as a data science evangelist promoting products via blogging and giving talks at meetups. Joe has a background in water engineering. Before his data science journey, he was an EngD research engineer at STREAM Industrial Doctorate Centre working on machine learning techniques for drainage design optimization. Prior to that, he was an asset management consultant specialized in data mining and constrained optimization for the utilities sector in UK and abroad. He also holds a MSc in Environmental Management and a BEng in Civil Engineering. At the end we will have some snacks. We want to thank H2O for their sponsorship.

  • Data …(don’t wait for it)… .table! (data.table Practical Guide)

    Carlos Bort and Jordi Puigdellivol will present how to process data using the data.table package. data.table allows to read and manipulate large datasets at very high speed thanks to its memory efficient treatment of data. Carlos and Jordi will present how to get started using data.table and guide us from basic to more advanced uses. Important: this will be a "hands on" session, bringing your laptop is recommended! When RSVP you'll be asked to give us some advice on what you you'd like to see in the next events and how to improve the meetings. Answer, attend the meetup and get the chance to win a price offered by RStudio! After the meetup some of the members usually hang out for a drink and a bite, which is great to get to know better fellow R users! You will find us at King's offices (Avinguda Josep Tarradellas 123) on the 6th floor. Please note that per King's policy you will have to register as a visitor at the entrance and that the audience is limited to 70 people. We'd be very grateful if you could RSVP to the meetup using your name and surname (to speedup the registration process) and cancel your attendance if you won't be able to join. See you at the meetup!

  • What's new in Spark for R/Advanced R: R6 Classes

    King Offices

    What's new in Spark for R (Krishna Kalyan) The aim of this talk will be to discuss whats new in SparkR 2.0.x. We will see how to write user defined functions using dapply, gapply, and lapply. I will also introduce machine learning, here we will learn how sparkR can be used to tune hyper-parameters, make predictions and save this model. Audience can bring their laptops if they want to follow along with the code ( All code/slides will be posted on Github before the meetup). I would like to present this talk for 30 mins and take questions at the end for another 5-10 minutes. Advanced R: From Environments through Classes to Reactive Programming We are going to introduce few advanced concepts like: * R environments, how to implement object oriented programming in R. * S3 classes, generic functions, double dispatch. * R6 clases and how they interact with S3 objects. * Implementing reactive programming (shiny's style) using environments/classes. * More examples from popular R packages.

  • 11th Intelligent Data Processing Conference 2016

    Cibernarium MediaTIC (Barcelona Activa)

    11th Intelligent Data Processing Conference 2016 Started in 1989, Intelligent Data Processing is one of the oldest and largest big data conferences in the world. The conference has it roots in computer vision but eventually grew to cover a broader scope of topics: - Computer Vision - Data Mining - Machine Learning - Big Data Analytics - Deep Learning - Text Mining - Social Networks Analysis - Other Data Science Applications Historically the conference focused on cutting edge research. As we move into 2016 we are transforming the event to have equal emphasis on both theoretical advances and real world impact that can benefit millions of people. We have separate but equally important research and industry tracks. IDP 2016 is the leading conference focusing on how data, machine learning, deep learning, computer vision and analytics are changing not only business, but society itself. We invite leaders from the industry and academia to present at the conference and foster exchange of ideas. The idea is to bring data scientists working on research together with those working in the field to accelerate adoption and testing of new approaches and techniques. The goal is to encourage ideas, success stories and challenges sharing among big data research and industry communities, present the latest technology advancements, connect with experienced data scientists in our community. As a result the conference will generate not only cutting edge research but also new products, features and technologies for companies, and new research problems for our academic colleagues. We will have some amazing speakers and panelists: Boris Polyak Institute for Control Science, Moscow Victor Lempitsky Skolkovo Institute of Science and Technology Konstantin Vorontsov Moscow Institute of Physics and Technology Amir Youssefi PayPal Sina Sohangir Snapchat Alexander Isakov Pallantius Konstantin Mertsalov Rational Enterprise Kamran Elahian 500S/GCP Christian Palau Sanz Red Arbor Pere Vallés Scytl John Clippinger MIT Media Lab Sima Yazdani Cisco Leonid Zhukov BCG Sebastian Wieczorek SAP Krishna Kashyap Accenture Pouya Tafti Allianz Pierre Gutierrez Dataiku Nikolay Zhukov Russian Railways Yahya Tabesh Stanford University Stephen Boyd Stanford University We would love to see everyone there!

  • Bayesian modelling in R via JAGS (and Bugs and Stan)

    The session will consist of a brief introduction to the logic of Bayesian inference, followed by a hands-on tutorial on how to specify Bayesian models in JAGS/Bugs/Stan via R, and how to do post-processing of MCMC models using tools for assessing convergence and for interpretation. Examples of actual uses of Bayesian inference will be presented. This talk will be done in english. Questions accepted in catalan, spanish and english. Xavier Fernández i Marín is senior researcher at ESADE Business School. He works on methodology of the social sciences, mainly in applications related to International Relations, Public Administration and Public Policy. He has created ggmcmc, a suite of graphical tools for analyzing Markov Chain Monte Carlo simulations from Bayesian inference. Xavier's personal webpage (http://xavier-fim.net/). Sponsor

  • Time Series Analysis: a Box-Jenkins R approach

    Social Point

    Box-Jenkins method is a a well known methodology in the field of Time Series Analysis presenting a framework to create ARIMA models in an structured way. In this session we will go through this framework together and learn how we can implement it using R. This talk will be done in english. Questions accepted in catalan, spanish and english. Iñaki Puigdollers Born and raise in Barcelona, with a mixed background in computer science and mathematics. Currently working as a Data Scientist at Social Point. Eager to learn and share knowledge. Update[9-3-16]: Notice we switched floors. We will meet on the 4th floor. Iñaki has prepared the talk as a kind of workshop. If you want to try it out, bring your laptop!!! You should have the latest version of R, RStudio and also car, fractal and tseries packages. Scrtipt can be downloaded from this link (http://files.meetup.com/10507982/Tseries%20demo%20RUGBCN.R). Sponsor:

  • Crear mapas con Leaflet + Montar RStudio en Amazon

    Crear mapas con Leaflet Leaflet (http://leafletjs.com/) es una librería JavaScript para mapas interactivos. Es utilizado por el The New York Times y The Washington Post a GitHub y Flickr, OpenStreetMap, Mapbox y CartoDB, entre otros. El equipo de RStudio ha creado un package con el mismo nombre para integrar de forma fácil mapas de Leaflet desde R (https://rstudio.github.io/leaflet/), usando la potencia de htmlwidgets (http://www.htmlwidgets.org/) . En este tutorial se enseñaran las principales funcionalidades que aporta este paquete y en que casos es más útil. François Delaunay Trabaja en los campos de la micro-meteorología y de la energía eólica desde 2001. Considera los datos meteorológicos como una forma de mirar a la madre naturaleza. Para esta afición suele usar R. Montar RStudio Server en Amazon ¿Tienes más datos de los que puedes tratar con tu PC/Mac? ¿Tienes prisa? Amazon Web Services (https://aws.amazon.com/es/) permite disponer de los servidores con tanta RAM como necesites a un precio asequible. En este tutorial, se mostrará como montar un RStudio server en Amazon de forma muy rápida y casi sin tener conocimientos de servidores, conexiones, etc como es mi caso. Esto es posible gracias a la imagen y el tutorial creado por Louis Aslett (http://www.louisaslett.com/RStudio_AMI/). Lluís Ramon (https://es.linkedin.com/in/lluisramon) Trabaja en el campo de la movilidad y la seguridad vial desde 2007. Considera el R una herramienta clave para su trabajo diario y para hacer el friki, sin el R es más difícil hehe. Las sesiones se realizaran en castellano, aunque se podrá preguntar en catalan, inglés o francés. UPDATE (27/1/16): Desde King nos informan que la sala es en la 9a planta. También nos comentan que os tendréis que registrar a una hoja de asistencia. Agradeceremos vuestra colaboración.

  • SparkR + H2O

    UPC Campus nord, AULARI A3 - AULA A3002 - anfiteatre

    In this special occasion the Barcelona Spark meetup (http://www.meetup.com/es/Spark-Barcelona/), Barcelona R Users Group (http://www.meetup.com/es/RugBcn-Barcelona-R-users-group/) and the Barcelona Machine Learning Study Group (http://www.meetup.com/es/Grup-destudi-de-machine-learning-de-Barcelona/), within the Big Data Week (http://barcelona.bigdataweek.com/) event, organize together a session about the combination of Spark, R and machine learning. In data management on a large scale we face with a problem of handling data and then treat them fast. For this reason we try to work in a distributed way. In this workshop we introduce the following tools that can help us. - SparkR: joins the distributed and robust processing, data sources, off memory data structures from Spark with the dynamic enviroment, interactivity, packages, and visualization tools from R. - H20: framework where its speed and flexibility allow users to fit hundreds or thousands of potential models as part of discovering patterns in data. We recommend bringing you computer with the software installed. Soon, we will send the instructions for the installation. Jordi Puigdellivol Freixa: Life Learner, Mathematician, Data Scientist & addicted to solving problems. I did a Bachelor's Degree in Mathematics at the Universitat Autònoma deBarcelona (UAB), where I discovered my passion for algorithms and artificial intelligence. I have also worked on business intelligence and as a consultant analyst. I'm currently working at Aia in data analysis for financial applications. Bartek Skorulski: Lifelong learner, mathematician by education, software developer by hobby and board/card/video game player. He finished his PhD in Mathematics in Dynamical Systems on Warsaw University of Technology. He had been working on various universities for several years and then he decided to try his luck outside the academic enviroment. Now he is working as Data Scientist at King. Maria José Peláez Montalvo: Data analyst at Schibsted and Mathematician specialized in Numeric Linear Algebra. She finished her undergraduate studies on University Complutense of Madrid and then she did PhD on Carlos III University of Madrid. Then for a long time she had been working on university doing research and teaching. Now she enjoys a lot working with this amazing area that is analysis of data. She likes to play with numbers that have people behind. Installation instructions: For those of you that want to follow the session with your own computer and do the exercises, please intall previously the required packages. SparkR: follow the instructions in the link http://sbartek.github.io/sparkRInstall/installSparkReasyWay.html H2O: first install the latest version of java, http://www.java.com/es/download/manual.jsp Then execute the following code in your R (or RStudio) console. #Uninstall previous version. if ("package:h2o" %in% search()) { detach("package:h2o", unload=TRUE) }if ("h2o" %in% rownames(installed.packages())) { remove.packages("h2o") } #needed packages!!!!#Curl needed.# needs libcurl-dev for RCurl## sudo apt-get install libcurl14-openssl-dev if (! ("methods" %in% rownames(installed.packages()))) { install.packages("methods") }if (! ("statmod" %in% rownames(installed.packages()))) { install.packages("statmod") }if (! ("stats" %in% rownames(installed.packages()))) { install.packages("stats") }if (! ("graphics" %in% rownames(installed.packages()))) { install.packages("graphics")}if (! ("RCurl" %in% rownames(installed.packages()))) { install.packages("RCurl") }if (! ("jsonlite" %in% rownames(installed.packages()))) { install.packages("jsonlite")}if (! ("tools" %in% rownames(installed.packages()))) { install.packages("tools") }if (! ("utils" %in% rownames(installed.packages()))) { install.packages("utils") } #install h2oinstall.packages("h2o", type="source", repos=(c(" http://h2o-release.s3.amazonaws.com/h2o/rel-tibshirani/3/R "))) #install h2o.Ensambledevtools::install_github("h2oai/h2o-2/R/ensemble/h2oEnsemble-package")