Note: this trajectory was previously called "Explainable Data Analytics"!
The fields of Process Mining and Visual Analytics are a unique specialization offered at TU Eindhoven. Both fields develop novel techniques for making data and models highly understandable through exploration and visualization. The courses in this trajectory provide you with solid foundations and advanced knowledge in both areas.
The courses in this trajectory teach you techniques to explore and integrate various kinds of complex data by designing problem-specific interactive visual analytics pipelines for data and model exploration, feature engineering, hypothesis generation and evaluation (2AMV10). You will learn how to analyze data from more than one viewpoint, revealing new features and properties that are crucial for better data understanding, and learning behavioral models (2AMV10, 2AMI20). You will learn how to construct highly explainable graph-based behavioral models and process models that we callglass-box models through unsupervised learning that visually describe any kind of behavior (processes) in an easily understandable manner (2AMI10, 2AMI20). On the example of process models, yYou will learn to adopt a mindset of evaluating models not just in terms of aggregate errors measures but along various quality dimensions including precise diagnostic information that can explain the quality of the model for each data point (2AMV10, 2AMI10, 2AMI20), and how to use the diagnostic information for improving model quality (2AMI20, 2AMV10). The course 2AMU30 Uncertainty Representation and Reasoning from the “AI and Machine Learning” trajectory covers this topic from the angle of Machine Learning and Artificial Intelligence.