arrangeres av Valeriya Naumova og resten av fagfolkene fra faggruppen BI & Analytics
The BI&Analytics professional group cordially invites you to the next meetup on AutoML and Low-code ML. The meetup will be held on April 05, 17:00-19:00.
AutoML is a set of tools that can automate the time consuming, iterative tasks of machine learning model development such as data preparation, feature engineering, model selection, optimization, and evaluation. It allows analysts, data scientists and developers to build ML models with high scale and efficiency.
In this meetup, we will learn what is AutoML, what type of features are available in state-of-the-art AutoML solutions; how AutoML / low-code solutions can be used for BI, and what is the data scientists view on the AutoML solutions. We gathered representatives from the technology provider and hands-on data scientists to facilitate the lively discussion on the topic.
17:05 – 17:30 Speaker: Christopher Frenning, Microsoft
Christopher has 20+ years experience in the software industry, with a background as developer and architect. He has experience with product development, leading development teams, and scaling a software startup to an international industry leader. He is passionate about the enhanced innovation made possible by cloud platforms, services, and technology. Christopher is leading the data center strategy for Microsoft Norway, and when not working he spends his time with his family, brews espresso, servicing more than riding bikes, and baking sourdough bread.
Automated machine learning aims at helping data scientists, analysts, and developers build ML models faster by automating some of the time-consuming, iterative tasks of model development. During training, AutoML creates a number of pipelines in parallel that try different algorithms and parameters for you. It scores each model, iterating to produce the best possible “fit” for your data. This talk gives an overview of Azure AutoML and its features, existing and upcoming, and a short demonstration of the low-code tools available in Azure Machine Learning Studio.
17:30-17:55 Speaker: Guro Larssen, Genus
Guro leads the «Consumer Goods» team in Genus – the leading low-code platform and community in the Nordics. She has extensive experience with business intelligence, process improvements and application development within leading enterprises like NorgesGruppen, Møller Mobility Group, DNB and SalMar.
What are the advantages of using low-code tools when working with BI? How does low-code tools help unlock a new range of possibilities when working with data insight? With the help of use cases from customers as NorgesGruppen among others, Guro will show how to add business value by connecting analytics to actions, while maintaining data integrity and security.
18:00 – 18:25 Speaker: Magnus Axelsson, Simula Consulting
Magnus is an experienced data science lead, having worked with many of Norway’s most well known companies such as FINN, Lendo, Tine, Equinor, etc. He has a PhD in Astrophysics, and has used statistical data analysis techniques to answer fundamental scientific questions for 15+ years. He thoroughly enjoys applying ML techniques to solve business problems, and works with all aspects of data science, and at one point built his own «AutoML» library. He currently works for Simula Consulting applying state-of-the-art machine learning research to solve challenging industry problems.
In this talk I will provide a data scientist’s perspective on low/no-code tools. Where in the data science life cycle could they be helpful, and for which specific tasks. On the other hand I will mention what limitations I see as an experienced user; that will lead to a few thoughts and speculations on how these tools perhaps could be improved.
18:30 – 19:00 – Panel discussion with the speakers
We will challenge our speakers and the audience with a number of interesting topics like why AutoML is important, whether it can replace data scientist, what use cases/industries that could benefit from AutoML in a near future.
Language: English and Norwegian