Integration of Data Mining Classification Techniques and Ensemble Learning for Predicting the Export Potential of a Company
| Autor: | Jenny Romero Borré, Ligia Castro, Aurora Patricia Piñeres Castillo, Noel Varela, Jesús Silva |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
competitiveness
Artificial neural network Computer science k-means clustering Decision tree 020206 networking & telecommunications data mining 02 engineering and technology computer.software_genre Ensemble learning Support vector machine Naive Bayes classifier ComputingMethodologies_PATTERNRECOGNITION K-Means clustering classification models 0202 electrical engineering electronic engineering information engineering export potential General Earth and Planetary Sciences 020201 artificial intelligence & image processing Data mining Cluster analysis computer General Environmental Science Test data |
| Zdroj: | ANT/EDI40 REDICUC-Repositorio CUC Corporación Universidad de la Costa instacron:Corporación Universidad de la Costa |
| ISSN: | 1877-0509 |
| DOI: | 10.1016/j.procs.2019.04.171 |
| Popis: | In this research, data mining techniques are integrated with Ensemble Learning for predicting the export potential of a company. The analysis covers the stages of measurement, evaluation and classification of companies, based on a proposal of 16 key factors of the export potential. The techniques standing out are: Synthetic Minority Oversampling Technique (Smote), K-Means Clustering, Generalized Regression Neural Network (GRNN), Feed Forward Back Propagation Neural Network (FFBPN), Support Vector Machine (SVM), Decision Tree (DT) and Naive Bayes. The neural network classifiers like GRNN and FFBPN are used for classification in MATLAB in the numeric form of data with a training and testing data ratio of 70% and 30% respectively. The accuracy of other classifiers such as SVM, DT and Naive Bayes is calculated on the nominal form of data with 80% data split. Artificial neural networks showed 85.7% of ability to discriminate and classify companies according to their competitive profile. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|