It's AI Time! With SMS group.

Dear members of our community, we are happy to announce our first on-site meetup in a long time!
We start with a special event in cooperation with SMS Group that will take place at Messe Düsseldorf on June 12th.
(You don't have to buy a ticket for the Exhibition to come.)

There will be three talks. First, we will hear from SMS group’s Maximilian Christ on “Explainability of time series models”.Later on, our special guest from the UK, Dr. Fawaz Ghali from Hazelcast will present the topic of real-time anomaly detection. Last but not least, Dr. Cedric Schockaert, the Head of Data Science at Paul Wurth will talk about Democratizing Machine Learning with a No-Code Platform.

Please find the detailed abstracts below.

Logistics:

• Address: GPS coordinates: 51.269011, 6.727094
  The booth is at Hall 1, E40/41.
  Address for car navigation system: D-40474 Düsseldorf, Am Staad (Stockumer Höfe)

• We recommend you to come by train: To get to the exhibition centre, take tram no. U78 MERKUR SPIEL-ARENA/Messe Nord or U79 Messe Ost (Exit at MesseOst/Stockumer Kirchstr.) or bus no. 722 (Exit at Messe Ost or Messe Süd/CCD). There are two major transport hubs where you are likely to change: Hauptbahnhof (central station) and Heinrich-Heine-Allee. Nearly all destinations in and around Düsseldorf can be reached from these stations.

• All listed Meetup participants are asked to register at the Info Counter at the North Entrance and receive a service ticket there. This service ticket entitles admission from 6 p.m. onwards. We will meet you there and lead you to the Meetup space.

Abstracts:

Speaker 1: Maximilian Christ
Lead Data Scientist, SMS digital

Title: Explainability of time series models
Abstract: In recent years, deep learning models have been increasingly used to analyze time series data in various domains, including the steel industry. However, these models are often regarded as black boxes, making it difficult to interpret and understand their predictions. This lack of explainability is a significant barrier to wider adoption, as it limits the ability of domain experts to make informed decisions based on the model's output.
The aim of this presentation is to discuss techniques for explaining time series deep learning models and breaking open the black box. We will illustrate the relevance of these techniques in the context of the steel industry, where the ability to identify countermeasures based on deep learning models is crucial for improving efficiency and reducing costs. By breaking open the black box, we can improve the transparency and interpretability of deep learning models, enabling domain experts to make more informed decisions and drive better outcomes in the steel industry and beyond.

Speaker 2: Fawaz Ghali, Dr. Principal Data Scientist

Title: Beyond Logs: Real-Time Anomaly Detection with Machine Learning
Abstract: Logs and traces generated by applications are valuable sources of information that can help detect issues and improve performance. However, they are often treated separately from other data, even though they are no different from the data an application works with. In this talk, we will explore a different approach: treating logs and traces as part of a scalable cloud storage repository that can be analyzed with the same techniques used for big data.

By keeping all the data together, we can apply machine learning models to detect situations of interest and alert us in real-time when unwanted behavior is occurring or brewing. This approach enables intelligent monitoring that goes beyond simple threshold-based alerts and can help identify complex issues that would otherwise go unnoticed. We will discuss how to harness existing technologies to implement this approach, providing attendees with practical tips and insights that they can apply to their own projects.

Speaker 3: Cedric Schockaert, Dr. Ir. Head of Data Science, Paul Wurth

Title: AIXpert: Democratizing Machine Learning with a No-Code Platform

Abstract: The democratization of machine learning has become increasingly essential as its applications continue to grow in various industries. However, the traditional process of training machine learning models is often hindered by the requirement of advanced programming skills, limiting its accessibility for the domain expert. In this talk, we will introduce AIXpert, a no-code platform designed to bridge this gap and empower users of all backgrounds to train machine learning models seamlessly without any coding experience.

AIXpert offers a user-friendly, intuitive interface that enables users to easily navigate through the process of model training, from data ingestion to model deployment. We will discuss the platform's innovative features, including automatic data preprocessing, model selection, and hyperparameter optimization, AutoML, which simplify the model training process while maintaining high-performance levels.