From schwendingerf at gmail.com Sat Apr 18 20:04:01 2020 From: schwendingerf at gmail.com (Florian Schwendinger) Date: Sat, 18 Apr 2020 20:04:01 +0200 Subject: [Viennar-meetup] Vienna<-R 2020 April Message-ID: Dear ViennaR members, Dienstag, 21. April 2020 18:30 bis 20:00 MESZ Online event Due to COVID-19 we offer our first online Meetup through Zoom and Youtube streaming. Youtube streaming: https://youtu.be/3r9C7XBrp6U Zoom: https://zoom.us/j/406260864 The Meetup is hosted together with Accenture Austria and two exciting speakers and topics. TOPICS 1. Dr. Wasif Masood, Magenta Telekom - Data Science Lead: Smart Churn Retention. (30min+Q/A) LinkedIn: https://www.linkedin.com/in/wasifmasood 2. Felix Glaser, Accenture - Data Scientist: Anomality Detection – an End To End Process with various R tools and packages (30min+Q/A) LinkedIn: https://www.linkedin.com/in/felix-glaser-a81325158 DETAILS 1. Dr. Masood describes one of Magenta's use cases built to reduce customer churn. He will explain the neural network model developed along with the baysian optimization that runs on top to find the best hyper-parameters. Additionally, the concept of shaply values is explained which was used to identify the reasons for churn. After finishing his PhD in information technology, Dr. Masood joined Magenta Telekom as a data scientist in 2017 and now heads the team as lead data scientist. His interests includes deep neural networks, ensemble trees and baysian optimization techniques. During his employment, he has worked on several use cases including network quality estimation, customer churn and retention, house hold matching, customer lifetime value, propensity modelling for x/up-selling, price to demand analysis, portfolio optimization and many more. 2. Data Anomalies are a hot topic for many clients spread over all industries. Detecting them is a hot topic for data scientists. As it often comes to the question “What are we actually looking for?”, a first step is towards unsupervised learning. With mathematical methods like Robust Principle Component Analysis we managed to distinguish anomalies from “usual” data points. Furthermore the detection tool was integrated in a webservice also built with R packages and was lifted into production. Felix is working for Accenture since 2018 as a Data Scientist. During his mathematics studies he specialized on forecasting, time series analysis and statistics, which also increased his R-Developer Skills. Knowledge in this topics made Felix participating in several machine learning projects at Accenture in public sector as well as financial sector. R as coding language was a central skill in many of those projects. Stay at Home and Healthy! Greetings, ViennaR organizers -------------- next part -------------- An HTML attachment was scrubbed... URL: