|
| This Article | ||
| ||
| Share | ||
| Bibliographic References | ||
| Add to: | ||
 Digg  Furl  Spurl  Blink  Simpy  Google  Del.icio.us  Y!MyWeb | ||
| Search | ||
Leave-one-out procedures for nonparametric error estimates
April 1989 (vol. 11 no. 4)
pp. 421,422,423
| ASCII Text | x | ||
| "Leave-one-out procedures for nonparametric error estimates," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 11, no. 4, pp. 421,422,423, April, 1989. | |||
| BibTex | x | ||
| @article{ 10.1109/34.19039, author = {}, title = {Leave-one-out procedures for nonparametric error estimates}, journal ={IEEE Transactions on Pattern Analysis and Machine Intelligence}, volume = {11}, number = {4}, issn = {0162-8828}, year = {1989}, pages = {421,422,423}, doi = {http://doi.ieeecomputersociety.org/10.1109/34.19039}, publisher = {IEEE Computer Society}, address = {Los Alamitos, CA, USA}, } | |||
| RefWorks Procite/RefMan/Endnote | x | ||
| TY - JOUR JO - IEEE Transactions on Pattern Analysis and Machine Intelligence TI - Leave-one-out procedures for nonparametric error estimates IS - 4 SN - 0162-8828 SP EP EPD - 421,422,423 PY - 1989 KW - estimation theory KW - Bayes methods KW - error analysis KW - nearest neighbor KW - leave one out procedures KW - nonparametric error estimates KW - covariances KW - Parzen estimates KW - Error analysis KW - Kernel KW - Testing KW - Covariance matrix KW - Nearest neighbor searches KW - Upper bound KW - Pattern recognition KW - Smoothing methods KW - Equations VL - 11 JA - IEEE Transactions on Pattern Analysis and Machine Intelligence ER - | |||
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/34.19039
The use of nonparametric error estimates may lead to biased results if the kernel covariances are estimated from the same data as are used to form the error estimate. If additional design samples are available, one may eliminate this bias by estimating the class covariances using an independent set of data. If, however, additional samples are not available, one may resort to leave-one-out type estimates of the kernel (for Parzen estimates) or metric (for nearest-neighbor estimates) for every sample being tested. The authors present an efficient algorithm for computation of these leave-one-out type estimates that requires little additional computational burden over procedures currently in use. The presentation is applicable to both Parzen and k-nearest neighbor (k-NN) type estimates. Experimental results demonstrating the efficiency of the algorithm are provided.<>
[1] K. Fukunaga and D. Hummels, "Bayes error estimation using Parzen andk-NN procedure,"IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI- 9, pp. 634-643, 1987.
[2] K. Fukunaga,Introduction to Statistical Pattern Recognition. New York: Academic, 1972.
Index Terms:
estimation theory,Bayes methods,error analysis,nearest neighbor,leave one out procedures,nonparametric error estimates,covariances,Parzen estimates,Error analysis,Kernel,Testing,Covariance matrix,Nearest neighbor searches,Upper bound,Pattern recognition,Smoothing methods,Equations
Citation:
"Leave-one-out procedures for nonparametric error estimates," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 11, no. 4, pp. 421,422,423, April 1989, doi:10.1109/34.19039
Usage of this product signifies your acceptance of the Terms of Use.