The lunch starts at 12:00, and the talk will start around 12:20.

NB: Wednesday BigInsight Lunches are open to staff and students from any of the BigInsight partners, including UiO, but not to others.

Speakers: Håkon Treider, Are Haartveit and Harald Husum from Intelecy AS.

Location: 8. floor Niels Henrik Abel’s building.

Title: Industrial machine learning with neural networks: Multivariate, unlabelled and noisy data

Abstract:

Process and manufacturing industry often possess vast amounts of data concerning their operations. This is time series data in the form of sensor data, event and alarm logs, but also less structured data like work orders, maintenance logs and product quality measurements. The data is typically stored for years while being under-utilised. As data scientists we see great potential here.

We think that with proper analysis, we can help make production cleaner, safer and more efficient. Data science and machine learning are usually sold to businesses as bespoke solutions that are expensive and time consuming. Intelecy wants to change this by providing off-the-shelf machine learning products for common industrial applications.

In this talk we discuss solutions for automatically categorising and preprocessing time series that lack metadata. Thereafter we explore challenges in making neural networks learn complex long-term patterns in event-driven sequences, as well as some recent approaches to mitigate these challenges.