Variable Ranking Feature Selection for the Identification of Nucleosome Related Sequences

Giosue' Lo Bosco, Alfonso Urso, Riccardo Rizzo, Giosué Lo Bosco, Alfonso Urso, Riccardo Rizzo, Massimo La Rosa, Antonino Fiannaca

Risultato della ricerca: Paper

Abstract

Several recent works have shown that K-mer sequence representation of a DNA sequence can be used for classification or identification of nucleosome positioning related sequences. This representation can be computationally expensive when k grows, making the complexity in spaces of exponential dimension. This issue affects significantly the classification task computed by a general machine learning algorithm used for the purpose of sequence classification. In this paper, we investigate the advantage offered by the so-called Variable Ranking Feature Selection method to select the most informative k − mers associated to a set of DNA sequences, for the final purpose of nucleosome/linker classification by a deep learning network. Results computed on three public datasets show the effectiveness of the adopted feature selection method.
Lingua originaleEnglish
Stato di pubblicazionePublished - 2018

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Mathematics(all)

Cita questo

Variable Ranking Feature Selection for the Identification of Nucleosome Related Sequences. / Lo Bosco, Giosue'; Urso, Alfonso; Rizzo, Riccardo; Lo Bosco, Giosué; Urso, Alfonso; Rizzo, Riccardo; La Rosa, Massimo; Fiannaca, Antonino.

2018.

Risultato della ricerca: Paper

Lo Bosco, Giosue' ; Urso, Alfonso ; Rizzo, Riccardo ; Lo Bosco, Giosué ; Urso, Alfonso ; Rizzo, Riccardo ; La Rosa, Massimo ; Fiannaca, Antonino. / Variable Ranking Feature Selection for the Identification of Nucleosome Related Sequences.
@conference{c70d8e775cfa4112a995196cbab331e0,
title = "Variable Ranking Feature Selection for the Identification of Nucleosome Related Sequences",
abstract = "Several recent works have shown that K-mer sequence representation of a DNA sequence can be used for classification or identification of nucleosome positioning related sequences. This representation can be computationally expensive when k grows, making the complexity in spaces of exponential dimension. This issue affects significantly the classification task computed by a general machine learning algorithm used for the purpose of sequence classification. In this paper, we investigate the advantage offered by the so-called Variable Ranking Feature Selection method to select the most informative k − mers associated to a set of DNA sequences, for the final purpose of nucleosome/linker classification by a deep learning network. Results computed on three public datasets show the effectiveness of the adopted feature selection method.",
author = "{Lo Bosco}, Giosue' and Alfonso Urso and Riccardo Rizzo and {Lo Bosco}, Giosu{\'e} and Alfonso Urso and Riccardo Rizzo and {La Rosa}, Massimo and Antonino Fiannaca",
year = "2018",
language = "English",

}

TY - CONF

T1 - Variable Ranking Feature Selection for the Identification of Nucleosome Related Sequences

AU - Lo Bosco, Giosue'

AU - Urso, Alfonso

AU - Rizzo, Riccardo

AU - Lo Bosco, Giosué

AU - Urso, Alfonso

AU - Rizzo, Riccardo

AU - La Rosa, Massimo

AU - Fiannaca, Antonino

PY - 2018

Y1 - 2018

N2 - Several recent works have shown that K-mer sequence representation of a DNA sequence can be used for classification or identification of nucleosome positioning related sequences. This representation can be computationally expensive when k grows, making the complexity in spaces of exponential dimension. This issue affects significantly the classification task computed by a general machine learning algorithm used for the purpose of sequence classification. In this paper, we investigate the advantage offered by the so-called Variable Ranking Feature Selection method to select the most informative k − mers associated to a set of DNA sequences, for the final purpose of nucleosome/linker classification by a deep learning network. Results computed on three public datasets show the effectiveness of the adopted feature selection method.

AB - Several recent works have shown that K-mer sequence representation of a DNA sequence can be used for classification or identification of nucleosome positioning related sequences. This representation can be computationally expensive when k grows, making the complexity in spaces of exponential dimension. This issue affects significantly the classification task computed by a general machine learning algorithm used for the purpose of sequence classification. In this paper, we investigate the advantage offered by the so-called Variable Ranking Feature Selection method to select the most informative k − mers associated to a set of DNA sequences, for the final purpose of nucleosome/linker classification by a deep learning network. Results computed on three public datasets show the effectiveness of the adopted feature selection method.

UR - http://hdl.handle.net/10447/295999

UR - https://link.springer.com/chapter/10.1007/978-3-030-00063-9_30

M3 - Paper

ER -