The grand tour: a tool for viewing multidimensional data
SIAM Journal on Scientific and Statistical Computing
Learning in the presence of concept drift and hidden contexts
Machine Learning
Nonlinear time series analysis
Nonlinear time series analysis
Hierarchical parallel coordinates for exploration of large datasets
VIS '99 Proceedings of the conference on Visualization '99: celebrating ten years
Data mining: concepts and techniques
Data mining: concepts and techniques
Mining time-changing data streams
Proceedings of the seventh ACM SIGKDD international conference on Knowledge discovery and data mining
Mining data streams under block evolution
ACM SIGKDD Explorations Newsletter
Semantic anomaly detection in online data sources
Proceedings of the 24th International Conference on Software Engineering
High-Performance Commercial Data Mining: A Multistrategy Machine Learning Application
Data Mining and Knowledge Discovery
High Dimensional Brushing for Interactive Exploration of Multivariate Data
VIS '95 Proceedings of the 6th conference on Visualization '95
Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
Neighborhood Property--Based Pattern Selection for Support Vector Machines
Neural Computation
High-dimensional data visualisation: The textile plot
Computational Statistics & Data Analysis
Info-fuzzy algorithms for mining dynamic data streams
Applied Soft Computing
Efficient mining of emerging events in a dynamic spatiotemporal environment
PAKDD'06 Proceedings of the 10th Pacific-Asia conference on Advances in Knowledge Discovery and Data Mining
| Hi-index | 0.00 |
We describe a visualization technique that uses brushed, parallel histograms to aid in understanding concept drift in multidimensional problem spaces. This technique illustrates the relationship between changes in distributions of multiple antecedent feature values and the outcome distribution. We can also observe effects on the relative utilization of predictive rules. Our parallel histogram technique solves the over-plotting difficulty of parallel coordinate graphs and the difficulty of comparing distributions of brushed and original data. We demonstrate our technique's usefulness in understanding concept drifts in power demand and stock investment returns.