Efficient Nearest Neighbor Classification Using a Cascade of Approximate Similarity Measures

Authors:
Vassilis Athitsos;Jonathan Alon;Stan Sclaroff
Affiliations:
Boston University;Boston University;Boston University
Venue:
CVPR '05 Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) - Volume 1 - Volume 01
Year:
2005

Citing 0
Cited 9

Query-sensitive embeddings

ACM Transactions on Database Systems (TODS)
BoostMap: An Embedding Method for Efficient Nearest Neighbor Retrieval

IEEE Transactions on Pattern Analysis and Machine Intelligence
IKNN: Informative K-Nearest Neighbor Pattern Classification

PKDD 2007 Proceedings of the 11th European conference on Principles and Practice of Knowledge Discovery in Databases
Use of Structured Pattern Representations for Combining Classifiers

SSPR & SPR '08 Proceedings of the 2008 Joint IAPR International Workshop on Structural, Syntactic, and Statistical Pattern Recognition
Investigating a novel GA-based feature selection method using improved KNN classifiers

International Journal of Information and Communication Technology
Learning pairwise image similarities for multi-classification using Kernel Regression Trees

Pattern Recognition
A real-time transportation prediction system

IEA/AIE'12 Proceedings of the 25th international conference on Industrial Engineering and Other Applications of Applied Intelligent Systems: advanced research in applied artificial intelligence
A real-time transportation prediction system

Applied Intelligence
Multi-label automatic indexing of music by cascade classifiers

Web Intelligence and Agent Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper proposes a method for efficient nearest neighbor classification in non-Euclidean spaces with computationally expensive similarity/distance measures. Efficient approximations of such measures are obtained using the BoostMap algorithm, which produces embeddings into a real vector space. A modification to the BoostMap algorithm is proposed, which uses an optimization cost that is more appropriate when our goal is classification accuracy as opposed to nearest neighbor retrieval accuracy. Using the modified algorithm, multiple approximate nearest neighbor classifiers are obtained, that provide a wide range of trade-offs between accuracy and efficiency. The approximations are automatically combined to form a cascade classifier, which applies the slower and more accurate approximations only to the hardest cases. The proposed method is experimentally evaluated in the domain of handwritten digit recognition using shape context matching. Results on theMNIST database indicate that a speed-up of two to three orders of magnitude is gained over brute force search, with minimal losses in classification accuracy.