Algorithms, Complex Networks, Time Series
Work description
The research work to be carried out will involve the following steps: Analyze existing work in interval time series forecasting, graph-based time series representations (e.g., quantile and visibility graphs), and network link predictions techniques for directed and weighted networks, and privacy-preserving collaborative forecasting approaches; Adapt or develop a link prediction approach suitable for forecasting on quantile graphs, considering directionality and edge weights to capture and predict time-dependent dynamics; Adapt or develop the inverse quantile graph process to reconstruct synthetic interval time series from updated (predicted) quantile graphs; Integrate the above methods into a collaborative forecasting pipeline, allowing multiple parties to contribute via graph representations without exchanging original time series data; Test the framework on real-world datasets, such as electricity consumption time series, and compare forecasting performance and privacy preservation against existing (possible) state-of-the-art methods; Compile findings into a final report.
Academic Qualifications
Bachelor degree in Computer Science, Computer Engineering, Data Science, or a related field.
Minimum profile required
Familiarity with methods for mapping time series to graphs.Ability to work independently, critical thinking, and interest in scientific research.
Preference factors
Knowledge and experience in complex network analysis and graph theory and time series analysis. Previous work related to the topics of the project. Experience with the C and Python programming languages.
Application Period
Since 26 Jun 2025 to 09 Jul 2025
Centre
Advanced Computing Systems