Type of Document	Master's Thesis
Author	Rishiyur, Adityan V
URN	etd-07142006-093118
Title	Instantaneously Trained Neural Networks with Complex and Quaternion Inputs
Degree	Master of Science (M.S.)
Department	Electrical & Computer Engineering
Advisory Committee	Advisor Name Title Subhash Kak Committee Chair Ashok Srivastava Committee Member Hsiao-Chun Wu Committee Member
Keywords	artificial intelligence image understanding Kak neural network
Date of Defense	2006-05-15
Availability	unrestricted
Abstract Neural network architectures such as backpropagation networks, perceptrons or generalized Hopfield networks can handle complex inputs but they require a large amount of time and resources for the training process. This thesis investigates instantaneously trained feedforward neural networks that can handle complex and quaternion inputs. The performance of the basic algorithm has been analyzed and shown how it provides a plausible model of human perception and understanding of images. The motivation for studying quaternion inputs is their use in representing spatial rotations that find applications in computer graphics, robotics, global navigation, computer vision and the spatial orientation of instruments. The problem of efficient mapping of data in quaternion neural networks is examined. Some problems that need to be addressed before quaternion neural networks find applications are identified.
Files	Filename       Size       Approximate Download Time (Hours:Minutes:Seconds)     28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access    Rishiyur_thesis.pdf 339.79 Kb 00:01:34 00:00:48 00:00:42 00:00:21 00:00:01