Statistical Signal Analysis Using Wavelets. Year 1 Report,
Abstract
This report discusses the research on "Statistical Signal Analysis Using Wavelets" performed by the Statistical Sciences Division of MathSoft, Inc. for contract N00014- 93-C-0106 with the Office of Naval Research. The overall goals of the research are: Application of wavelets and related transforms to data analysis; Using wavelets as the basis for statistical problems, such as signal extraction, spectral density estimation, isotonic regression, and classification; Research into new transforms and algorithms tailored to meet the needs of data analysis and statistical estimation. The first year of research focused on six specific areas: (1) Applications of wavelets to data of interest to the Navy; (2) Noise removal using wavelets, wavelet packets, and cosine packets; (3) Investigation into new wavelet transforms which are outlier resistant and edge preserving; (4) Development of a framework and tools for the "wavelet approach" towards analysis of signals, images, and other data; (5) Exploration of the use of wavelets as a dimension reduction tool for statistical analysis of very large data sets; (6) Development of algorithms for wavelet analysis. AN
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 17, 1995
- Accession Number
- ADA298917
Entities
People
- R. D. Martin