Skip to Main content Skip to Navigation
Theses

Classification des séquences d’ADN par les réseaux d’ondelettes

Abstract : now days, there has been an increasing interest in the research of the classification of the sequences of the DNA. However, despite the multiplicity at the level of methods and strategies used, a radical solution has been far from being obtained. In this study, we have tried to provide a contribution to the theoretical study of wavelet neural network and the theoretical study of the classification of the sequences of the DNA. We have developed a new approach. This approach is inspired by wavelet neural networks Beta. This network is built through a new method which combines a gradual algorithm and a predictor called the Least Trimmed Squares(LTS) or a selection method by orthogonalisation .We have relied on a new approach called Power Spectrum to identify the distribution of nucleotides at the level of DNA sequence. The evaluation of the approach used has been made through the use of many DNA data bases. To measure the degree of the efficiency of our model, we have applied many criteria of evaluation, such as, Precision, Recall and confusion matrix.
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-01771864
Contributor : Abdesselem Dakhli <>
Submitted on : Thursday, April 19, 2018 - 11:55:54 PM
Last modification on : Thursday, April 26, 2018 - 11:28:49 AM
Long-term archiving on: : Tuesday, September 18, 2018 - 3:05:43 PM

File

These.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-01771864, version 1

Citation

Abdesselem Dakhli. Classification des séquences d’ADN par les réseaux d’ondelettes . Informatique [cs]. ENIS-SFAX, 2017. Français. ⟨tel-01771864⟩

Share

Metrics

Record views

157

Files downloads

231