Given an encoded unknown text message in the form of a three dimensional spatial series generated by the use of four smooth nonlinear functions, we use a method based on simple statistical reasoning to pick up samples for rebuilding the four functions. The estimated functions are then used to decode the sequence. The experimental results show that our method gives a nearly perfect decoding, enabling us to submit a 100% accurate solution to the IJCNN challenge problem
X. Cai et al., "A Statistical Solution to a Text Decoding Challenge Problem," Proceedings of the International Joint Conference on Neural Networks, 2001., Institute of Electrical and Electronics Engineers (IEEE), Jan 2001.
The definitive version is available at http://dx.doi.org/10.1109/IJCNN.2001.939504
International Joint Conference on Neural Networks, 2001.
Electrical and Computer Engineering
Keywords and Phrases
IJCNN Challenge Problem; Binary Sequences; Decoding; Encoded Unknown Text Message; Nearly Perfect Decoding; Nonlinear Functions; Smooth Nonlinear Functions; Statistical Analysis; Statistical Reasoning; Statistical Solution; Text Decoding; Three Dimensional Spatial Series
Article - Conference proceedings
© 2001 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.