Download presentation
Presentation is loading. Please wait.
1
Interactive Visualization of the Stock Market Graph Presented by Camilo Rostoker rostokec@cs.ubc.ca Department of Computer Science University of British Columbia
2
Stock Market Data Stock market produces huge amounts of data on a daily basis, and its easy to acquire Stock market data consists of a variety of fields such as price, volume, change, change %, etc. Stock market data Take samples of stock data at regular intervals for a large set of stocks Convert dataset to correlation matrix correlation(x,y) = [-1, +1]
3
The Market Graph Convert correlation matrix to a graph, where Vertices represent stocks edge(x,y) correlation(x,y) >= threshold High threshold few edges Low threshold more edges The market graph has been shown to have small world properties
4
What Are We Visualizing? Maximum Cliques Highly positively/negatively correlated subsets of stocks Independent Sets Completely diversified stocks Quasi-Cliques/Independent Sets Generalizations allow for near matches Find clusters/groups of stocks that exhibit certain trading patterns
5
Usage Scenarios Portfolio management (static) Real-time market analysis (dynamic) Exploratory analysis of trading data to gain new insights, spot patterns/trends, etc (static)
6
Implementation Extend H3 – Hyperbolic 3D browser Rational: Good focus+context view supports interactive data exploration Convenient API for interactive control and navigation of graphs Stock market graph is large hyperbolic space has good information density
7
Adapting & Extending H3 Colour-encode clusters Encode inter-cluster links thickness, colour Create “dummy nodes” to represent clusters encode aggregrate info Keyboard controls for basic interaction Dynamic graph capabilities Click interaction for information integration
8
Current Prototype
9
References Vladimir Boginski, Sergiy Butenko, and Panos M. Pardalos. Mining market data: A network approach. Tamara Munzner. H3: Laying out large directed graphs in 3d hyperbolic space. In Proceedings of the 1997 IEEE Symposium on Information Visualization, pages 2-10, 1997. James Chilson, Raymond Ng, Alan Wagner, and Ruben Zamar. Parallel computation of high dimensional robust correlation and covariance matrices. In KDD 04: Proceedings of the 2004 ACM SIGKDD international conference on Knowledge discovery and data mining, pages 533-538, New York, NY, USA, 2004. ACM Press. Wayne Pullan. Phased local search. Journal TBA. 2005.
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.