Automatic view selection: an application to image mining

Authors:
Manoranjan Dash;Deepak Kolippakkam
Affiliations:
School of Computer Engineering, Nanyang Technological University, Singapore;Department of Computer Science and Engg., Arizona State University, Tempe, AZ
Venue:
PAKDD'05 Proceedings of the 9th Pacific-Asia conference on Advances in Knowledge Discovery and Data Mining
Year:
2005

Citing 3
Cited 0

Selection of relevant features and examples in machine learning

Artificial Intelligence - Special issue on relevance
Mining frequent patterns without candidate generation

SIGMOD '00 Proceedings of the 2000 ACM SIGMOD international conference on Management of data
Adaptive View Validation: A First Step Towards Automatic View Detection

ICML '02 Proceedings of the Nineteenth International Conference on Machine Learning

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we discuss an image mining application of Egeria detection. Egeria is a type of weed found in various lands and water regions over San Joaquin and Sacramento deltas. The challenge is to find a view to accurately detect the weeds in new images. Our solution contributes two new aspects to image mining. (1) Application of view selection to image mining: View selection is appropriate when a specific learning task is to be learned. For example, to look for an object in a set of images, it is useful to select the appropriate views (a view is a set of features and their assigned values). (2) Automatic view selection for accurate detection: Usually classification problems rely on user-defined views. But in this work we use association rule mining to automatically select the best view. Results show that the selected view outperforms other views including the full view.