|
Project Summary
The objective of this activity is to develop a general Knowledge Discovery (KD) tool for handling Large Data Sets, to support scientific exploratory data analysis, during and after any mission.With advances of computer power and Technologies we are now able to collect and store considerable large amounts of data. However, existing tools to explore, analyze and visualize these large data sets are rather insufficient and quite limited! We need to dramatically improve scalability on ordinary and/or parallel computers to handle Large and Complex Data Sets and discover new information/knowledge. Examples of astronomical missions that collect considerable amounts of data, yet barely studied in-depth, are: Sloan Digital Sky Survey, which stores around 53 million unique objects, Hipparcos, which stores around 10^5 unique objects and GAIA future mission, which will collect 10^9 unique objects with 10% being variable objects (10^8). This tool is targeted for astronomers, physicists, mission control teams and other interested parties, because it will allow exploration of data during and after the missions, to obtain more knowledge about the Universe. The aim of KD-LADS is to allow an in-depth knowledge-discovery process, complemented with specialized visualizations techniques (e.g. SOM- Self Organizing Maps) to better visualize the data clusters and provide more insights for exploring and analyzing the data, thus improving any knowledge discovery process.
Research Areas
- Knowledge Discovery
- Data Mining
- Parallel computing
Partners Publications Simas, T.; Silva, G.; Miranda, B.; Moitinho, A.; Ribeiro, R., Knowledge Discovery in Large Data Sets, CLASSIFICATION AND DISCOVERY IN LARGE ASTRONOMICAL SURVEYS: Proceedings of the International Conference: “Classification and Discovery in Large Astronomical Surveys”. AIP Conference Proceedings, Volume 1082, pp. 196-200 (2008).
Links
|
