Designing efficient cascaded classifiers: tradeoff between accuracy and cost

Authors:
Vikas C. Raykar;Balaji Krishnapuram;Shipeng Yu
Affiliations:
Siemens Healthcare, Malvern, PA, USA;Siemens Healthcare, Malvern, PA, USA;Siemens Healthcare, Malvern, PA, USA
Venue:
Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining
Year:
2010

Citing 6
Cited 1

Text Categorization Based on Regularized Linear Classification Methods

Information Retrieval
Robust Object Detection via Soft Cascade

CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 2 - Volume 02
An Expected Utility Approach to Active Feature-Value Acquisition

ICDM '05 Proceedings of the Fifth IEEE International Conference on Data Mining
Cost-sensitive feature acquisition and classification

Pattern Recognition
Partial example acquisition in cost-sensitive learning

Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining
LungCAD: a clinically approved, machine learning system for lung cancer detection

Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining

Using classifier cascades for scalable e-mail classification

Proceedings of the 8th Annual Collaboration, Electronic messaging, Anti-Abuse and Spam Conference

Quantified Score

Hi-index	0.00

Visualization

Abstract

We propose a method to train a cascade of classifiers by simultaneously optimizing all its stages. The approach relies on the idea of optimizing soft cascades. In particular, instead of optimizing a deterministic hard cascade, we optimize a stochastic soft cascade where each stage accepts or rejects samples according to a probability distribution induced by the previous stage-specific classifier. The overall system accuracy is maximized while explicitly controlling the expected cost for feature acquisition. Experimental results on three clinically relevant problems show the effectiveness of our proposed approach in achieving the desired tradeoff between accuracy and feature acquisition cost.