EARL 2018, London
Conference Day 1 {.tabset .tabset-fade}
Wednesday 12 September
###Keynote
####Edwina Dunn, Starcount
####Garrett Grolemund, RStudio
###Session 1 ####1. “A Validated R Environment in the Cloud for Life Science R&D” Jobst Loffler, Bayer Business Services GmbH
- Waiting on Rstudio
- Item 2
####2. “A brief history of Data at Autotrader; how R has got us here” Paul Owens, Autotrader
- Item 1
- Item 2
####3. “Making multi-million pound baseball decisions with R, Shiny and H2O AutoML” Jo-Fai Chow, H2O.ai
- Item 1
- Item 2
###Session 2 ####1. “Interpretable Machine Learning with LIME – now and tomorrow” Kasia Kulma, Aviva
####2. “Costing the Armed Forces using the Tidyverse” Tomas Westlake, Ministry of Defence
####2. “Forecasting Work Demand with Random Forests” Eduardo Contreras Cortes, Ernst & Young
###Lightning Talks
##
Conference Day 2 {.tabset .tabset-fade}
Thursday 13 September
###Keynote ####Richard Pugh, Chief Data Scientist, Mango Solutions
- Community in Practice
- Identity
- Communication
- Technology
- Hackathon
- Internal Events
- Best Practices
- Analytic Governace
- Objectives, Definitions, Prioristations, Language, Deploy, Quality, Measure, Innovation, Ethics
- Community of Practice + Best Practices + Analytic Gonverance = A consistent and coherent analytic capabiltiy focused on delivering value via high quality analytics
- know how to buld the “engine” … now it needs to drive the car!
- Big Data Anallytics Ecosystem = A datalake of all available data + tools to access & analyse data + capable analysts to provide insight
- strategic ideas and
- Educate the business
- Language (analytics, use cases and results), Preconceptions, Inspire
- Focus on Decisions
- Seek Enthusiasm, Workshops, Decisions
- Measure Success
- Agree Metrics, Measure, Report, Evangelise
Panel
- boring
Session1
####1. “Data Visualisation in R Shiny: from Interactive reports to Automated Tools” Taisiya Merkulova, Photobox
- “By visualizing information, we turn it into a landscape that you can explore with your eyes, a sort of information map. And when your’re lost in information, an information map is kind of useful.” - David McCandless, Data journalist
- build interactive web apps, free visualization tool, effective, versatile, one-stop-shop
- collect & store, extract, process, visualize
- snapshots in between to improve performance
####2. “Re-think not re-write: bridging quant modelling and engineering with R” Scott Finnie and Nick Forrester, Hymans Robertson
Relevant packages:
- pacman
- flexdashboard
- googleAnalyticsR
- googleAuthR
- quanteda
- here
####3. “Shiny Graph Networks for Organisation Design” Jonathan Ng, HSBCY
- “Cheat Codes to Unlick the Power of Graphs & Level Up Your Analytics Skillset”
- Website
- DataStrategyWithjonathan.com/p/shiny
- https://datastrategywithjonathan.com/courses/enrolled/364839
- MS Visio
- graphs are not just for visulizations
- Data Access Layer; Business Logic Layer; Presentation Layer
- library(RNeo4j), cypher
- igraph
- tidygraph
- regression
- flexdashboard
- vis.js
- https://github.com/ukgovdatascience/orgsurveyr
Session2
####1. “The power of machine learning in segmenting CRM databases” Jeremy Horne, MC&C Media
- CRM analytics
####2. “Future of R in our enterprise (and perhaps yours)” Alexis Iglauer, PartnerRe
- Insurance
- gitlab
- focus on a single analytics platform; centralised infrastructure; hiring to an absolute bar
- more communication; more collaborations
- dedicated R-Studio und Shiny server
- Docker & cloud are the future
####3. “Hello Soda Bridging the gap between Data Scientists and Engineers; using R in production” Leanne Fitzpatrick
- Barrier Spider
- Docker
- Plumber
- Schemas with YAML
- pacman
- cookie cutter (reproducable framework)
- testthat & usethis
Session3
####1. “Training 3,000 R models to predict parcel volumes using Docker and Microsoft Azure” Christoph Bodner, Austrian Post
- Booh!
- why? https://about.gitlab.com/product/project-management/
- how? https://stanford.edu/~shervine/teaching/cs-229/cheatsheet-machine-learning-tips-and-tricks#classification-metrics
####2. “A ‘caret’-based Framework for Training Multiple Tax Fraud Detection Models” Lars Kjeldgaard, Danish Tax Authority
##