|
Showing 1 - 10 of
10 matches in All Departments
A classic problem in the field of pattern recognition is that of
handwritten digit recognition. The heart of the problem lies within
the ability to design an efficient algorithm that can recognize
digits written and submitted by users via a tablet, scanner, and
other digital devices. Previous studies have shown that the digits
can be modeled as points in a high dimensional space which can
adequately approximated by linear structures. In Chapters 2 and 3,
we present two previously established algorithms rooted in two
distinct geometric frameworks. In Chapter 4, we introduce a novel
approach to the problem based on the linear structure that arises
for the Grassmann manifold. From this framework, we developed three
separate algorithms. We then tested our algorithms on the publicly
available MNIST database and benchmarked our results with the
algorithm described in Chapters 2 and 3. Without any preprocessing
of the images, one of the Grassmann-based algorithms simultaneously
achieved the highest classification rate and one of the fastest
processing times.
This is a reproduction of a book published before 1923. This book
may have occasional imperfections such as missing or blurred pages,
poor pictures, errant marks, etc. that were either part of the
original artifact, or were introduced by the scanning process. We
believe this work is culturally important, and despite the
imperfections, have elected to bring it back into print as part of
our continuing commitment to the preservation of printed works
worldwide. We appreciate your understanding of the imperfections in
the preservation process, and hope you enjoy this valuable book.
|
|