https://oldena.lpnu.ua/handle/ntb/40244
Title: | Handwritten characters recognition using modification of classic Rosenblatt’s perceptron extended with multiple S-layers |
Authors: | Serdyuk, Pavlo Mosiiuk, Vasyl |
Affiliation: | Lviv Polytechnic National University |
Bibliographic description (Ukraine): | Serdyuk P. Handwritten characters recognition using modification of classic Rosenblatt’s perceptron extended with multiple S-layers / Pavlo Serdyuk, Vasyl Mosiiuk // Litteris et Artibus : proceedings of the 6th International youth science forum, November 24–26, 2016, Lviv, Ukraine / Lviv Polytechnic National University. – Lviv : Lviv Polytechnic Publishing House, 2016. – P. 41–42. – Bibliography: 6 titles. |
Conference/Event: | Litteris et Artibus |
Issue Date: | 2016 |
Publisher: | Lviv Polytechnic Publishing House |
Country (code): | UA |
Place of the edition/event: | Lviv |
Keywords: | computer vision CV perceptron OCR machine learning handwritten characters |
Number of pages: | 41-42 |
Abstract: | Presented structure and principles of perceptron with multiple sensor layers. Demonstrated how it recognize character image and vote for result. |
URI: | https://ena.lpnu.ua/handle/ntb/40244 |
References (International): | [1] [F. Rosenblatt, "Principles of Neurodynamic" Spartan Books, Washington, DC, 1962. [2] [M. L. Minsky, "Perceptrons" MIT Press, Cambridge, MA, 1969. [3] [S. I. Gallant, "Perceptron-based learning algorithms" IEEE Transactions on Neural Networks, vol. 1, no. 2, pp. 179–191, 1990. [4] [Y. Hongfeng, "Perceptron-Based Algorithms and Analysis" Spectrum Library, Concordia University, Canada, 1996. [5] M. D. Estebon, "Perceptrons: An Associative Learning Network" Virginia Tech, 1997. [6] J. E. Dayhoff, " Neural Network Architecturs: An introduction" Van Nostrand Reinhold, Australia, 19970. |
Content type: | Conference Abstract |
Appears in Collections: | Litteris et Artibus. – 2016 р. |
File | Description | Size | Format | |
---|---|---|---|---|
LEA-2016-41-42.pdf | 415.41 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.