The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning

Yaosan Lin, Liangsian Lin, Der-Chiang Li, Hungyu Chen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In the past few decades, there were quite a few learning algorithms developed to extract knowledge from data. However, none of the single algorithms can be applicable to learn all the datasets with favor results because data patterns may represent linear and non-linear. Accordingly, the idea of aggregating the predictions of multiple learning models to improve the forecasting accuracy of a single method was proposed. Nevertheless, how to improve the accuracy of the aggregated predictions when learning small datasets is the objective of this study. Based on the distributions of the predictive errors of learning models, the proposed method learns the weights of the models and then tries to aggregate more precise predictions with the weights. The experiment results show the forecasting errors of the predictions aggregated by the proposed method are significantly lower than the predictions of single models and the averaged predictions.

Original languageEnglish
Title of host publicationProceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages79-85
Number of pages7
ISBN (Electronic)9781538638866
DOIs
Publication statusPublished - 2018 Mar 28
Event2017 International Conference on Computing Intelligence and Information System, CIIS 2017 - Nanjing, Jiangsu, China
Duration: 2017 Apr 212017 Apr 23

Publication series

NameProceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017
Volume2018-January

Other

Other2017 International Conference on Computing Intelligence and Information System, CIIS 2017
CountryChina
CityNanjing, Jiangsu
Period17-04-2117-04-23

Fingerprint

Multi-model
Prediction
Forecasting
Model
Learning
Learning Algorithm
Learning algorithms
Experiment

All Science Journal Classification (ASJC) codes

  • Hardware and Architecture
  • Information Systems
  • Control and Optimization
  • Artificial Intelligence
  • Computer Networks and Communications

Cite this

Lin, Y., Lin, L., Li, D-C., & Chen, H. (2018). The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning. In Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017 (pp. 79-85). (Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017; Vol. 2018-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CIIS.2017.21
Lin, Yaosan ; Lin, Liangsian ; Li, Der-Chiang ; Chen, Hungyu. / The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning. Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 79-85 (Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017).
@inproceedings{f82f969c197e4492950a8fe7a84c3f48,
title = "The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning",
abstract = "In the past few decades, there were quite a few learning algorithms developed to extract knowledge from data. However, none of the single algorithms can be applicable to learn all the datasets with favor results because data patterns may represent linear and non-linear. Accordingly, the idea of aggregating the predictions of multiple learning models to improve the forecasting accuracy of a single method was proposed. Nevertheless, how to improve the accuracy of the aggregated predictions when learning small datasets is the objective of this study. Based on the distributions of the predictive errors of learning models, the proposed method learns the weights of the models and then tries to aggregate more precise predictions with the weights. The experiment results show the forecasting errors of the predictions aggregated by the proposed method are significantly lower than the predictions of single models and the averaged predictions.",
author = "Yaosan Lin and Liangsian Lin and Der-Chiang Li and Hungyu Chen",
year = "2018",
month = "3",
day = "28",
doi = "10.1109/CIIS.2017.21",
language = "English",
series = "Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "79--85",
booktitle = "Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017",
address = "United States",

}

Lin, Y, Lin, L, Li, D-C & Chen, H 2018, The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning. in Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017. Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017, vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 79-85, 2017 International Conference on Computing Intelligence and Information System, CIIS 2017, Nanjing, Jiangsu, China, 17-04-21. https://doi.org/10.1109/CIIS.2017.21

The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning. / Lin, Yaosan; Lin, Liangsian; Li, Der-Chiang; Chen, Hungyu.

Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017. Institute of Electrical and Electronics Engineers Inc., 2018. p. 79-85 (Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017; Vol. 2018-January).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning

AU - Lin, Yaosan

AU - Lin, Liangsian

AU - Li, Der-Chiang

AU - Chen, Hungyu

PY - 2018/3/28

Y1 - 2018/3/28

N2 - In the past few decades, there were quite a few learning algorithms developed to extract knowledge from data. However, none of the single algorithms can be applicable to learn all the datasets with favor results because data patterns may represent linear and non-linear. Accordingly, the idea of aggregating the predictions of multiple learning models to improve the forecasting accuracy of a single method was proposed. Nevertheless, how to improve the accuracy of the aggregated predictions when learning small datasets is the objective of this study. Based on the distributions of the predictive errors of learning models, the proposed method learns the weights of the models and then tries to aggregate more precise predictions with the weights. The experiment results show the forecasting errors of the predictions aggregated by the proposed method are significantly lower than the predictions of single models and the averaged predictions.

AB - In the past few decades, there were quite a few learning algorithms developed to extract knowledge from data. However, none of the single algorithms can be applicable to learn all the datasets with favor results because data patterns may represent linear and non-linear. Accordingly, the idea of aggregating the predictions of multiple learning models to improve the forecasting accuracy of a single method was proposed. Nevertheless, how to improve the accuracy of the aggregated predictions when learning small datasets is the objective of this study. Based on the distributions of the predictive errors of learning models, the proposed method learns the weights of the models and then tries to aggregate more precise predictions with the weights. The experiment results show the forecasting errors of the predictions aggregated by the proposed method are significantly lower than the predictions of single models and the averaged predictions.

UR - http://www.scopus.com/inward/record.url?scp=85050544107&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050544107&partnerID=8YFLogxK

U2 - 10.1109/CIIS.2017.21

DO - 10.1109/CIIS.2017.21

M3 - Conference contribution

T3 - Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017

SP - 79

EP - 85

BT - Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Lin Y, Lin L, Li D-C, Chen H. The Prediction Aggregating Procedure for Multi-models in Small Dataset Learning. In Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017. Institute of Electrical and Electronics Engineers Inc. 2018. p. 79-85. (Proceedings - 2017 International Conference on Computing Intelligence and Information System, CIIS 2017). https://doi.org/10.1109/CIIS.2017.21