The Strong Consistency of Error Probability Estimates in NN Discrimination.
Abstract
Since 1971, T.J. Wagner introduced an error probability estimate in NN (Nearest Neighbor) discrimination rule and proved the weak convergency of this estimate. The problem of strong consistency is not yet solved. This paper shows the exponential convergence rate of this estimate. Hence, the strong consistency of this estimate holds automatically. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1984
- Accession Number
- ADA160016
Entities
People
- Z. D. Bai
Organizations
- University of Pittsburgh