Date of Award
8-2009
Document Type
Thesis
Degree Name
Master of Science (MS)
Legacy Department
Electrical Engineering
Committee Chair/Advisor
Taha, Tarek M
Committee Member
Birchfield , Stanley M
Committee Member
Ligon , Walter B
Abstract
There have been considerable developments in the quest for intelligent machines since the beginning of the cybernetics revolution and the advent of computers. In the last two decades with the onset of the internet the developments have been extensive. This quest for building intelligent machines have led into research on the working of human brain, which has in turn led to the development of pattern recognition models which take inspiration in their structure and performance from biological neural networks. Research in creating intelligent systems poses two main problems. The first one is to develop algorithms which can generalize and predict accurately based on previous examples. The second one is to make these algorithms run fast enough to be able to do real time tasks. The aim of this thesis is to study and compare the accuracy and multi-core performance of some of the best learning algorithms to the task of handwritten character recognition. Seven algorithms are compared for their accuracy on the MNIST database, and the test set accuracy (generalization) for the different algorithms are compared. The second task is to implement and compare the performance of two of the hierarchical Bayesian based cortical algorithms, Hierarchical Temporal Memory (HTM) and Hierarchical Expectation Refinement Algorithm (HERA) on multi-core architectures. The results indicate that the HTM and HERA algorithms can make use of the parallelism in multi-core architectures.
Recommended Citation
Mohan, Sumod, "ACCURACY AND MULTI-CORE PERFORMANCE OF MACHINE LEARNING ALGORITHMS FOR HANDWRITTEN CHARACTER RECOGNITION" (2009). All Theses. 634.
https://tigerprints.clemson.edu/all_theses/634