Gradient-based optimization of hyperparameters for base-pairing profile local alignment kernels.

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

We have recently proposed novel kernel functions, called base-pairing profile local alignment (BPLA) kernels for discrimination and detection of functional RNA sequences using SVMs. We employ STRAL's scoring function which takes into account sequence similarities as well as upstream and downstream base-pairing probabilities, which enables us to model secondary structures of RNA sequences. In this paper, we develop a method for optimizing hyperparameters of BPLA kernels with respect to discrimination accuracy using a gradient-based optimization technique. Our experiments show that the proposed method can find a nearly optimal set of parameters much faster than the grid search on all parameter combinations.

Original languageEnglish
Pages (from-to)128-138
Number of pages11
JournalGenome informatics. International Conference on Genome Informatics
Volume23
Issue number1
Publication statusPublished - 2009 Oct
Externally publishedYes

Fingerprint

Base Pairing

ASJC Scopus subject areas

  • Medicine(all)

Cite this

@article{325d3342c2c54b5f87665bf931a93ba4,
title = "Gradient-based optimization of hyperparameters for base-pairing profile local alignment kernels.",
abstract = "We have recently proposed novel kernel functions, called base-pairing profile local alignment (BPLA) kernels for discrimination and detection of functional RNA sequences using SVMs. We employ STRAL's scoring function which takes into account sequence similarities as well as upstream and downstream base-pairing probabilities, which enables us to model secondary structures of RNA sequences. In this paper, we develop a method for optimizing hyperparameters of BPLA kernels with respect to discrimination accuracy using a gradient-based optimization technique. Our experiments show that the proposed method can find a nearly optimal set of parameters much faster than the grid search on all parameter combinations.",
author = "Kengo Sato and Yutaka Saito and Yasubumi Sakakibara",
year = "2009",
month = "10",
language = "English",
volume = "23",
pages = "128--138",
journal = "Genome informatics. International Conference on Genome Informatics",
issn = "0919-9454",
publisher = "Universal Academy Press",
number = "1",

}

TY - JOUR

T1 - Gradient-based optimization of hyperparameters for base-pairing profile local alignment kernels.

AU - Sato, Kengo

AU - Saito, Yutaka

AU - Sakakibara, Yasubumi

PY - 2009/10

Y1 - 2009/10

N2 - We have recently proposed novel kernel functions, called base-pairing profile local alignment (BPLA) kernels for discrimination and detection of functional RNA sequences using SVMs. We employ STRAL's scoring function which takes into account sequence similarities as well as upstream and downstream base-pairing probabilities, which enables us to model secondary structures of RNA sequences. In this paper, we develop a method for optimizing hyperparameters of BPLA kernels with respect to discrimination accuracy using a gradient-based optimization technique. Our experiments show that the proposed method can find a nearly optimal set of parameters much faster than the grid search on all parameter combinations.

AB - We have recently proposed novel kernel functions, called base-pairing profile local alignment (BPLA) kernels for discrimination and detection of functional RNA sequences using SVMs. We employ STRAL's scoring function which takes into account sequence similarities as well as upstream and downstream base-pairing probabilities, which enables us to model secondary structures of RNA sequences. In this paper, we develop a method for optimizing hyperparameters of BPLA kernels with respect to discrimination accuracy using a gradient-based optimization technique. Our experiments show that the proposed method can find a nearly optimal set of parameters much faster than the grid search on all parameter combinations.

UR - http://www.scopus.com/inward/record.url?scp=77952994541&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77952994541&partnerID=8YFLogxK

M3 - Article

VL - 23

SP - 128

EP - 138

JO - Genome informatics. International Conference on Genome Informatics

JF - Genome informatics. International Conference on Genome Informatics

SN - 0919-9454

IS - 1

ER -