⚠️ In order to be prepared for the workshop, we ask participants to go through the instructions in the document below.
Setup_Instructions.pdf ⚠️

In recent years, industry-wide trends like Digital Twins and Predictive Maintenance have led to significant interest in machine and deep learning methods for engineering applications. However, engineers still have some reservations towards these methods and can find it difficult to apply machine learning for their own projects. Furthermore, it is also becoming increasingly clear that applying domain-specific knowledge and methods for data preprocessing and feature engineering is a crucial component in developing a good machine learning model. 

In this workshop we will develop a machine learning model for condition monitoring of an air compressor based on recorded audio data to automatically detect different defects. 

We will explore different ways to extract relevant time- and frequency-domain features and perform feature selection to optimize the performance of our machine learning models. In a next step, we train a wide range of machine and deep learning models, explore the automatic selection and optimization of machine learning models, and discuss how to evaluate the performance of our models in order to pick the “best” solution. We will also talk about interpretability and explainability, and showcase some methods that can help us reveal how these black-box models make their prediction.

Participants will be provided access to a MATLAB Online environment with ready-made interactive notebooks and graphical apps so they can focus on the technical task at hand. The online environment will remain available for participants until one month after the workshop.

By the end of the workshop, people will be familiar with the main steps of the machine learning workflow and will have gained insight into techniques for feature extraction, training machine learning models and interpretability. 

Attendees will also have written MATLAB code and will have created an executable notebook which they can reuse and adapt for their own projects. 

• Basic understanding of machine learning concepts
• Programming experience (MATLAB experience is helpful but not required)
• Laptop with internet access and browser (Edge or Chrome recommended)

The hands-on parts of the workshop are done with MATLAB. You will have the choice to either use the MATLAB Online platform or the desktop version.

⚠️ In order to be prepared for the workshop, we ask participants to go through the instructions in the document below.
Setup_Instructions.pdf ⚠️