This year's Applied R Munich Spring Edition will be hosted & sponsored by CYOSS (location, food & drinks).
06:30 pm - Doors open
07:00 pm - Introduction CYOSS & First talk (see below) + Discussion
07:45 pm - Break
08:15 pm - Second talk (see below) + Discussion
There will be time for networking / food / drinks before, in between and after the talks.
Please, don't hesitate to contact us at [masked] (or via twitter https://twitter.com/applied_r) if your company wants to host one of our next meetups or if you have an interesting talk about an R related topic.
Talk 1: Sampling, Intervention, Prediction, Aggregation: A Generalized Framework for Model Agnostic Interpretations in Machine Learning – Theory and Demonstration
Bio of the speaker:
Christian Scholbeck is a Ph.D. candidate at the Department of Statistics at Ludwigs-Maximilians-University Munich. His research focuses on interpretable machine learning.
Non-linear machine learning models often trade off a great predictive performance for a lack of interpretability. However, model agnostic interpretation techniques now allow us to estimate the effect and importance of features for any predictive model. Different notations and terminology have complicated their understanding and how they are related. A unified view on these methods has been missing.
By deconstructing available methods into sequential work stages, one discovers striking similarities. We present the generalized SIPA (Sampling, Intervention, Prediction, Aggregation) framework of work stages for model agnostic interpretation techniques. All prominent model agnostic methods to interpret black box machine learning models such as Individual Conditional Expectation & Partial Dependence Plots, Accumulated Local Effects or the Permutation Feature Importance are based on the SIPA framework. The SIPA work stages served as an inspiration for the development of the R package iml (interpretable machine learning).
After an introduction to the theory of the SIPA framework and some methods that are based on it, we will gain insights into the workings of a trained black box machine learning model by using several methods from the R package iml.
Talk 2: Picking up the pieces – building a comprehensive open-source web platform for R Shiny Apps
Bio of the speaker:
Michael Mikesky is a Senior Data Scientist @ CYOSS
Using open source data Michael will guide through the process of building a open-source intelligence dashboard and implementing it into a web platform using Shinyproxy, Docker, Keycloak and other open-source technologies.
The presented system can be used for any R Shiny based application and will exemplarily demonstrate visualizations of country information from the World Bank related to poverty and prosperity sharing.
The key point in this framework is the combination of open-source technologies to create a comprehensive platform, which allows for user and role-based authentication and scalability of R Shiny applications.