### Abstract

Given a network of N nodes with the i-th sensor's observation χi ε R^{M}, the matrix containing all Euclidean distances among measurements ||χi - χj || ∀i, j ε {1, . . . , N} is a useful description of the data. While reconstructing a distance matrix has wide range of applications, we are particularly interested in the manifold reconstruction and its dimensionality reduction for data fusion and query. To make this map available to the all of the nodes in the network, we propose a fully decentralized consensus gossiping algorithm which is based on local neighbor communications, and does not require the existence of a central entity. The main advantage of our solution is that it is insensitive to changes in the network topology and it is fully decentralized. We describe the proposed algorithm in detail, study its complexity in terms of the number of inter-node radio transmissions and showcase its performance numerically.

Original language | English (US) |
---|---|

Title of host publication | ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings |

Pages | 3353-3356 |

Number of pages | 4 |

DOIs | |

State | Published - 2009 |

Externally published | Yes |

Event | 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009 - Taipei, Taiwan, Province of China Duration: Apr 19 2009 → Apr 24 2009 |

### Other

Other | 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009 |
---|---|

Country | Taiwan, Province of China |

City | Taipei |

Period | 4/19/09 → 4/24/09 |

### Fingerprint

### Keywords

- Dimensionality reduction
- Distributed computing
- Manifold estimation

### ASJC Scopus subject areas

- Signal Processing
- Software
- Electrical and Electronic Engineering

### Cite this

*ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings*(pp. 3353-3356). [4960343] https://doi.org/10.1109/ICASSP.2009.4960343

**Distributed distance estimation for manifold learning and dimensionality reduction.** / Yildiz, Mehmet E.; Ciaramello, Frank; Scaglione, Anna.

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

*ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings.*, 4960343, pp. 3353-3356, 2009 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2009, Taipei, Taiwan, Province of China, 4/19/09. https://doi.org/10.1109/ICASSP.2009.4960343

}

TY - GEN

T1 - Distributed distance estimation for manifold learning and dimensionality reduction

AU - Yildiz, Mehmet E.

AU - Ciaramello, Frank

AU - Scaglione, Anna

PY - 2009

Y1 - 2009

N2 - Given a network of N nodes with the i-th sensor's observation χi ε RM, the matrix containing all Euclidean distances among measurements ||χi - χj || ∀i, j ε {1, . . . , N} is a useful description of the data. While reconstructing a distance matrix has wide range of applications, we are particularly interested in the manifold reconstruction and its dimensionality reduction for data fusion and query. To make this map available to the all of the nodes in the network, we propose a fully decentralized consensus gossiping algorithm which is based on local neighbor communications, and does not require the existence of a central entity. The main advantage of our solution is that it is insensitive to changes in the network topology and it is fully decentralized. We describe the proposed algorithm in detail, study its complexity in terms of the number of inter-node radio transmissions and showcase its performance numerically.

AB - Given a network of N nodes with the i-th sensor's observation χi ε RM, the matrix containing all Euclidean distances among measurements ||χi - χj || ∀i, j ε {1, . . . , N} is a useful description of the data. While reconstructing a distance matrix has wide range of applications, we are particularly interested in the manifold reconstruction and its dimensionality reduction for data fusion and query. To make this map available to the all of the nodes in the network, we propose a fully decentralized consensus gossiping algorithm which is based on local neighbor communications, and does not require the existence of a central entity. The main advantage of our solution is that it is insensitive to changes in the network topology and it is fully decentralized. We describe the proposed algorithm in detail, study its complexity in terms of the number of inter-node radio transmissions and showcase its performance numerically.

KW - Dimensionality reduction

KW - Distributed computing

KW - Manifold estimation

UR - http://www.scopus.com/inward/record.url?scp=70349226729&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349226729&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2009.4960343

DO - 10.1109/ICASSP.2009.4960343

M3 - Conference contribution

AN - SCOPUS:70349226729

SN - 9781424423545

SP - 3353

EP - 3356

BT - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

ER -