Fast Multiscale Algorithms for Information Representation and Fusion
Abstract
In the third quarter of the work effort, we focused on the research and design of the randomized Approximate Nearest Neighbors (ANN) algorithm. This randomized variant of the ANN algorithm has theoretically proven improvements in the number of data dimensions that it can handle over existing algorithms and meets the theoretical lower bounds for computational complexity. Algorithm designs for computing the Randomized Approximate Nearest Neighbors (ANN) using randomized Fast Fourier Transform projections were completed. Fortran 95 interface for reusable randomized ANN routine has been defined and implemented. The randomized ANN implementation uses BLAS libraries via standardized interfaces to make optimal use of hardware resources (e.g., multiple cores, CPU cache) in addition to using the OpenMP standard (for parallel execution of code). Use of these standards enables the code to be built flexibly in a number of ways on various target platforms. Preliminary testing of the software is complete. Additional updating, fine tuning will be based on results from various experiments that will be conducted in the upcoming quarter. The project is currently on track - in the upcoming quarter, we will focus on testing and conducting experiments for the randomized SVD and ANN algorithms. This also includes documentation and packaging efforts. No problems are currently anticipated.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 2011
- Accession Number
- ADA543835
Entities
People
- Devasis Bassu