### Abstract

In this paper, we focus on pattern recognition based on the vector space model. As one of the methods, distance metric learning is known for the learning metric matrix under the arbitrary constraint. Generally, it uses iterative optimization procedure in order to gain suitable distance structure by considering the statistical characteristics of training data. Most of the distance metric learning methods estimate suitable metric matrix from all pairs of training data. However, the computational cost is considerable if the number of training data increases in this setting. To avoid this problem, we propose the way of learning distance metric by using the each category centroid. To verify the effectiveness of proposed method, we conduct the simulation experiment by using benchmark data.

Original language | English |
---|---|

Title of host publication | Proceedings - 2015 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015 |

Publisher | Institute of Electrical and Electronics Engineers Inc. |

Pages | 1645-1650 |

Number of pages | 6 |

ISBN (Print) | 9781479986965 |

DOIs | |

Publication status | Published - 2016 Jan 12 |

Event | IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015 - Kowloon Tong, Hong Kong Duration: 2015 Oct 9 → 2015 Oct 12 |

### Other

Other | IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015 |
---|---|

Country | Hong Kong |

City | Kowloon Tong |

Period | 15/10/9 → 15/10/12 |

### Fingerprint

### Keywords

- Distance Metric Learning
- Pattern Recognition
- Regularization
- Vector Space Model

### ASJC Scopus subject areas

- Artificial Intelligence
- Computer Networks and Communications
- Energy Engineering and Power Technology
- Information Systems and Management
- Control and Systems Engineering

### Cite this

*Proceedings - 2015 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015*(pp. 1645-1650). [7379422] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SMC.2015.290

**A Study of Distance Metric Learning by Considering the Distances between Category Centroids.** / Mikawa, Kenta; Kobayashi, Manabu; Goto, Masayuki; Hirasawa, Shigeichi.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

*Proceedings - 2015 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015.*, 7379422, Institute of Electrical and Electronics Engineers Inc., pp. 1645-1650, IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015, Kowloon Tong, Hong Kong, 15/10/9. https://doi.org/10.1109/SMC.2015.290

}

TY - GEN

T1 - A Study of Distance Metric Learning by Considering the Distances between Category Centroids

AU - Mikawa, Kenta

AU - Kobayashi, Manabu

AU - Goto, Masayuki

AU - Hirasawa, Shigeichi

PY - 2016/1/12

Y1 - 2016/1/12

N2 - In this paper, we focus on pattern recognition based on the vector space model. As one of the methods, distance metric learning is known for the learning metric matrix under the arbitrary constraint. Generally, it uses iterative optimization procedure in order to gain suitable distance structure by considering the statistical characteristics of training data. Most of the distance metric learning methods estimate suitable metric matrix from all pairs of training data. However, the computational cost is considerable if the number of training data increases in this setting. To avoid this problem, we propose the way of learning distance metric by using the each category centroid. To verify the effectiveness of proposed method, we conduct the simulation experiment by using benchmark data.

AB - In this paper, we focus on pattern recognition based on the vector space model. As one of the methods, distance metric learning is known for the learning metric matrix under the arbitrary constraint. Generally, it uses iterative optimization procedure in order to gain suitable distance structure by considering the statistical characteristics of training data. Most of the distance metric learning methods estimate suitable metric matrix from all pairs of training data. However, the computational cost is considerable if the number of training data increases in this setting. To avoid this problem, we propose the way of learning distance metric by using the each category centroid. To verify the effectiveness of proposed method, we conduct the simulation experiment by using benchmark data.

KW - Distance Metric Learning

KW - Pattern Recognition

KW - Regularization

KW - Vector Space Model

UR - http://www.scopus.com/inward/record.url?scp=84964514099&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84964514099&partnerID=8YFLogxK

U2 - 10.1109/SMC.2015.290

DO - 10.1109/SMC.2015.290

M3 - Conference contribution

AN - SCOPUS:84964514099

SN - 9781479986965

SP - 1645

EP - 1650

BT - Proceedings - 2015 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2015

PB - Institute of Electrical and Electronics Engineers Inc.

ER -