Edit metric decoding: a new hope

In this position paper we examine preliminary results of a new type of general error correction decoder for Edit Metric Codes. The Single Classifier Machine Decoder uses the concept of Side Effect Machines(SEMs) created via Genetic Algorithms(GAs) in order to create a mapping from the Edit Metric to the Euclidean Metric to create a decoder. By not having to measure the edit distance to every codeword the decoder has a far smaller runtime complexity. Fuzzy versions are also examined which reduce the number of times the Levenshtein or Edit distance must be calculated. Codes of the form (n, M, d)4 are targeted due to their suitability for bioinformatics problems. A (12, 55, 7)4 code is used as an example of the process.

[1]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[2]  Vladimir I. Levenshtein,et al.  Binary codes capable of correcting deletions, insertions, and reversals , 1965 .

[3]  D. Ashlock,et al.  Side effect machines for sequence classification , 2008, 2008 Canadian Conference on Electrical and Computer Engineering.

[4]  D. Ashlock,et al.  Construction of Optimal Edit Metric Codes , 2006, 2006 IEEE Information Theory Workshop - ITW '06 Chengdu.

[5]  Daniel A. Ashlock,et al.  Classifying synthetic and biological DNA sequences with side effect machines , 2008, 2008 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology.

[6]  Michael J. Fischer,et al.  The String-to-String Correction Problem , 1974, JACM.