CASUS Institute Seminar, Dr. Holger Brandl, analytics solution architect at SYSTEMA (Dresden, Germany)
At SYSTEMA, Holger’s work is focused around developing machine learning and cloud solutions for Industrial IoT and analytics. His commitment to revealing patterns in large unstructured data sets leads to quick and efficient delivery of actionable insights. The open-source tools, methods and algorithms he develops for factory optimization, high-performance computing, and data science are aligned with and supported by the latest technology and science. He holds a Ph.D. in machine learning, has developed novel concepts in computational linguistics, and recently co-authored systems biology publications in high-ranking journals including Nature and Science.
Here’s what Holger will present in his talk:
Kotlin’s language design and its great tooling provide a wonderful framework for data science. In this session, Holger will present an overview of the kotlin data science stack.
Starting with krangl, a {K}otlin DSL for data w{rangl}ing will be discussed. By mimicking well-established concepts from pandas and R, it implements a grammar of data manipulation using a modern functional-style API. It allows to filter, transform, aggregate, and reshape tabular data.
Next on the agenda is a discussion about how to use kotlin for scripting as an alternative to bash and python. To put some meat on this exciting topic, Holger will walk the audience through the main features of kscript, which is a common Kotlin script interpreter.
Finally, the talk will focus on an emerging library for the simulation of complex systems. kalasim is an open-source library written in Kotlin to enable discrete event simulation. To showcase this library, a simulation of a semiconductor frontend fab will be presented in conjunction with krangl, kscript, let-plot, and Jupyter.