I will first give a quick summary of Learning from Data Streams, and of Continual Learning, including some recent work on Online Continual Learning. I will give an overview of the TAIAO project, which stands for “Time-Evolving Data Science and Artificial Intelligence for Advanced Open Environmental Science”. Finally, I will quickly present the works of my current and recently finished PhD students, comprising the following topics:
- Advanced Adaptive Classifier Methods for Data Streams
- SO-KNL: Self-optimising K-Nearest Leaves Regression Ensembles
- Anomaly Detection in Streaming Data
- AutoML for Data Streams
- Self-supervised Feature Extractor Training for Alzheimer’s Disease Classification
- Feature Extractor Stacking for Cross-domain Few-shot Learning
- ML Approaches for Malware Classification based on Hybrid Artefacts
- Using LLMs to assess cybersecurity thread notes
- Fake News detection in Urdu
- Normalising Flows for Environmental Data
- Fast Clustering using GPUs
