Neural Network Ensembles to Determine Growth Multi-classes in Predictive Microbiology

Research areas:

• Methodology - Machine learning

Year:

2012

Type of Publication:

In Proceedings

Keywords:

Negative Correlation Learning, Neural Networks, Ordinal Regression

Authors:

• Fernandez-Navarro, Francisco
• Chen, Huanhuan
• Gutiérrez, Pedro Antonio
• Hervás-Martínez, César
• Yao, Xin

Book title:

7th International Conference on Hybrid Artificial Intelligence Systems (HAIS2012)

Pages:

308-318

BibTex:

@conference{NCL-Navarro2012, author = "Francisco Fernandez-Navarro and Huanhuan Chen and Pedro Antonio Guti{\'e}rrez and C{\'e}sar Herv{\'a}s-Mart{\'i}nez and Xin Yao", abstract = "This paper evaluates the performance of different ordinal regression, nominal classifiers and regression models when predicting probability growth of the Staphylococcus Aureus microorganism. The prediction problem has been formulated as an ordinal regression problem, where the different classes are associated to four values in an ordinal scale. The results obtained in this paper present the Negative Correlation Learning as the best tested model for this task. In addition, the use of the intrinsic ordering information of the problem is shown to improve model performance. ", booktitle = "7th International Conference on Hybrid Artificial Intelligence Systems (HAIS2012)", doi = "10.1007/978-3-642-28931-6_30", keywords = "Negative Correlation Learning, Neural Networks, Ordinal Regression", pages = "308-318", title = "{N}eural {N}etwork {E}nsembles to {D}etermine {G}rowth {M}ulti-classes in {P}redictive {M}icrobiology", url = "http://dx.doi.org/10.1007/978-3-642-28931-6_30", year = "2012", }

Abstract:

This paper evaluates the performance of different ordinal regression, nominal classifiers and regression models when predicting probability growth of the Staphylococcus Aureus microorganism. The prediction problem has been formulated as an ordinal regression problem, where the different classes are associated to four values in an ordinal scale. The results obtained in this paper present the Negative Correlation Learning as the best tested model for this task. In addition, the use of the intrinsic ordering information of the problem is shown to improve model performance.