Inbjudan till predoc-seminarium på Institutionen för data- och systemvetenskap (DSV) vid Stockholms universitet. Doktoranden Zed Lee presenterar sitt pågående arbete. Titeln är ”Z-Series: Mining and learning from complex sequential data”.

Respondent: Zed Lee, DSV
Opponent: Élisa Fromont, Université de Rennes, France
Huvudhandledare: Panagiotis Papapetrou, DSV
Handledare: Tony Lindgren, DSV
Närmast berörda professor: Hercules Dalianis, DSV

Sammanfattning på engelska

The rapid increase in the amount and complexity of sequential data collected in various domains has made it challenging to extract useful knowledge from such sources. This type of data can take different forms, such as multivariate time series, histogram snapshots, and heterogeneous health records, with varying granularity and multiple variables describing a single data instance.

Therefore, finding underlying temporal relations between those variables is crucial to understanding the underlying structure of these complex sequential data and capturing patterns of interest.

This thesis proposes using event intervals as a unified representation of different types of complex sequential data. The development of artifacts that exploit event intervals aims to efficiently and effectively capture patterns of interest, transform various sequential data in several application domains by applying temporal abstraction, and provide interpretable features for machine learning tasks.

Kontaktuppgifter till Zed Lee