### Abstract

The main feature of neural network using for accuracy improvement of physical quantities (for example, temperature, humidity, pressure etc.) measurement by data acquisition systems is insufficient volume of input data for predicting neural network training at an initial exploitation period of sensors. The authors have proposed the technique of data volume increasing for predicting neural network training using integration of historical data method. In this paper we have proposed enhanced integration historical data method with its simulation results on mathematical models of sensor drift using single-layer and multi-layer perceptrons. We also considered a parallelization technique of enhanced integration historical data method in order to decrease its working time. A modified coarse-grain parallel algorithm with dynamic mapping on processors of parallel computing system using neural network training time as mapping criterion is considered. Fulfilled experiments have showed that modified parallel algorithm is more efficient than basic parallel algorithm with dynamic mapping, which does not use any mapping criterion.

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

Title of host publication | Proceedings of the Third Workshop - 2005 IEEE Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005 |

Pages | 66-73 |

Number of pages | 8 |

DOIs | |

Publication status | Published - 2007 |

Event | 3rd IEEE Workshop on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005 - Sofia, Bulgaria Duration: Sep 5 2005 → Sep 7 2005 |

### Other

Other | 3rd IEEE Workshop on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005 |
---|---|

Country | Bulgaria |

City | Sofia |

Period | 9/5/05 → 9/7/05 |

### Fingerprint

### Keywords

- Coarse-grain parallel algorithm
- Computational grids
- Dynamic mapping
- Integration historical data
- Neural networks
- Sensor drift

### ASJC Scopus subject areas

- Artificial Intelligence
- Computer Vision and Pattern Recognition
- Theoretical Computer Science

### Cite this

*Proceedings of the Third Workshop - 2005 IEEE Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005*(pp. 66-73). [4062094] https://doi.org/10.1109/IDAACS.2005.282943

**Parallel algorithm of enhanced historical data integration using neural networks.** / Turchenko, V.; Triki, C.; Grandinetti, L.; Sachenko, A.

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

*Proceedings of the Third Workshop - 2005 IEEE Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005.*, 4062094, pp. 66-73, 3rd IEEE Workshop on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005, Sofia, Bulgaria, 9/5/05. https://doi.org/10.1109/IDAACS.2005.282943

}

TY - GEN

T1 - Parallel algorithm of enhanced historical data integration using neural networks

AU - Turchenko, V.

AU - Triki, C.

AU - Grandinetti, L.

AU - Sachenko, A.

PY - 2007

Y1 - 2007

N2 - The main feature of neural network using for accuracy improvement of physical quantities (for example, temperature, humidity, pressure etc.) measurement by data acquisition systems is insufficient volume of input data for predicting neural network training at an initial exploitation period of sensors. The authors have proposed the technique of data volume increasing for predicting neural network training using integration of historical data method. In this paper we have proposed enhanced integration historical data method with its simulation results on mathematical models of sensor drift using single-layer and multi-layer perceptrons. We also considered a parallelization technique of enhanced integration historical data method in order to decrease its working time. A modified coarse-grain parallel algorithm with dynamic mapping on processors of parallel computing system using neural network training time as mapping criterion is considered. Fulfilled experiments have showed that modified parallel algorithm is more efficient than basic parallel algorithm with dynamic mapping, which does not use any mapping criterion.

AB - The main feature of neural network using for accuracy improvement of physical quantities (for example, temperature, humidity, pressure etc.) measurement by data acquisition systems is insufficient volume of input data for predicting neural network training at an initial exploitation period of sensors. The authors have proposed the technique of data volume increasing for predicting neural network training using integration of historical data method. In this paper we have proposed enhanced integration historical data method with its simulation results on mathematical models of sensor drift using single-layer and multi-layer perceptrons. We also considered a parallelization technique of enhanced integration historical data method in order to decrease its working time. A modified coarse-grain parallel algorithm with dynamic mapping on processors of parallel computing system using neural network training time as mapping criterion is considered. Fulfilled experiments have showed that modified parallel algorithm is more efficient than basic parallel algorithm with dynamic mapping, which does not use any mapping criterion.

KW - Coarse-grain parallel algorithm

KW - Computational grids

KW - Dynamic mapping

KW - Integration historical data

KW - Neural networks

KW - Sensor drift

UR - http://www.scopus.com/inward/record.url?scp=43549086684&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=43549086684&partnerID=8YFLogxK

U2 - 10.1109/IDAACS.2005.282943

DO - 10.1109/IDAACS.2005.282943

M3 - Conference contribution

AN - SCOPUS:43549086684

SN - 0780394461

SN - 9780780394469

SP - 66

EP - 73

BT - Proceedings of the Third Workshop - 2005 IEEE Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications, IDAACS 2005

ER -