Evaluación del algoritmo AR-NSGEP en colecciones de datos desbalanceadas.

• Alain Guerrero Enamorado ^[1] ; Carlos Morell ^[2] ; Sebastián Ventura ^[3]
  1. [1] Universidad de las Ciencias Informáticas

     Universidad de las Ciencias Informáticas

     Cuba

     
  2. [2] Universidad Central de Las Villas

     Universidad Central de Las Villas

     Cuba

     
  3. [3] Universidad de Córdoba

     Universidad de Córdoba

     Cordoba, España

     

  Mostrar afiliaciones +
• Localización: Revista Cubana de Ciencias Informáticas, ISSN-e 2227-1899, Vol. 12, Nº. Extra 1, 2018 (Ejemplar dedicado a: Especial UCIENCIA)
• Idioma: español
• Títulos paralelos:
  • Evaluation of the AR-NSGEP algorithm in unbalanced datasets.
• Enlaces
  • Texto completo
• Resumen
  • español

    Uno de los grandes problemas que tiene la minería de datos es la existencia del desbalance. Este fenómeno puede afectar gravemente la efectividad de los sistemas de clasificación. Este trabajo persigue como objetivo fundamental obtener información empírica del desempeño del algoritmo AR-NSGEP en colecciones de datos no-balanceados. Se evaluá dicho algoritmo en colecciones de datos con diferentes niveles de desbalance. Se utilizaron colecciones con razones de desbalance entre 1,5 y 40. Durante la etapa de evaluación se utilizaron técnicas de validación cruzada y pruebas estadísticas no-paramétricas para consolidar los resultados obtenidos. La evaluación se realizó con tres métricas muy utilizadas para medir el desempeño en Sistemas Clasificadores con Aprendizaje. Los resultados obtenidos muestran la competitividad del algoritmo AR-NSGEP en colecciones de datos no-balanceados.

  • English

    One of the biggest problems with data mining is the existence of imbalance. This phenomenon can seriously affect the effectiveness of classification systems. The main objective of this work is to obtain empirical information of the performance of the AR-NSGEP algorithm in unbalanced datasets. This algorithm is evaluated in datasets with different levels of imbalance. Were used datasets with an unbalance rate between 1,5 and 40. During the evaluation stage, cross-validation techniques and non-parametric statistical tests were used to consolidate the results obtained. The evaluation was carried out with three metrics widely used to measure the performance in Learning Classifier Systems. The obtained results show the competitiveness of the AR-NSGEP algorithm in unbalanced data collections.

• Referencias bibliográficas
  • Bria, A,Karssemeijer, N,Tortorella, F. Learning from unbalanced data. A cascade-based ap- proach for detecting clustered microcalcifications.....
  • Calvo, B,Santafé, G. scmamp: Statistical Comparison of Multiple Algorithms in Multiple Problems.. 8. 248-256
  • Chawla, N. V,Bowyer, K. W,Hall, L. O,Kegelmeyer, W. P. SMOTE: Synthetic minority over-sampling technique.. 16. 321-357
  • Deb, K. (2014). Multi-objective Optimization.: Search Methodologies: Introductory Tutorials in Optimization and Decision Support Techniques....
  • Deb, K,Jain, H. An Evolutionary Many-Objective Optimization Algorithm Using Reference-point Based Non-dominated Sorting Approach. IEEE Transactions...
  • Deb, K,Pratap, A,Agarwal, S,Meyarivan, T. A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-II. IEEE Transactions on evolutionary...
  • Derrac, J,García, S,Molina, D,Herrera, F. A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing...
  • Dua, D,Karra Taniskidou, E. (2017). Machine Learning Repository.
  • Espejo, P. G,Ventura, S,Herrera, F. A Survey on the Application of Genetic Programming to Classification. 40. 121-144
  • Fernández, A,García, S,Bernadó-Mansilla, E,Herrera, F. (2009). Genetics-Based Machine Learning for Rule Induction: Taxonomy ,: Experimental...
  • Ferreira, C. (2006). Gene Expression Programming Mathematical Modeling by an Artificial Intelligence.. Springer. Berlin.
  • Gollapudi, S. (2016). Practical Machine Learning.: Tackle the real-world complexities of modern machine learning with innovative and cutting-edge...
  • Guerrero-Enamorado, A,Morell, C,Ventura, S. A gene expression programming algorithm for discovering classification rules in the multi-objective...
  • Holland, J. H,Reitman, J. O. Cognitive Systems Based on Adaptive Algorithms.. 63. 49-49
  • Japkowicz, N,S, Stephen. The class imbalance problem: A systematic study.. 6. 429
  • Sanatkar, H,Haratizadeh, S. An XCS-Based Algorithm for Classifying Imbalanced Datasets. 4. 101-105
  • Shen, X,Guo, Y,Chen, Q,Hu, W. A multi-objective optimization evolutionary algorithm incorporating preference information based on fuzzy logic.....
  • Triguero, I,González, S,Moyano, J. M,García, S,Alcalá-Fdez, J,Luengo, J,Fernández, A,Del Jeús, M,Sánchez, L,Herrera, F. KEEL 3.0: An Open...
  • R. J., Urbanowicz,Browne, W. N. (2017). Introduction to Learning Classifier Systems.: Springer Briefs in Intelligent Systems. Springer. Berlin....
  • Wang, S,Minku, L. L.,Yao, X. Resampling-Based Ensemble Methods for Online Class Imbalance Learning.. 27. 1356-1368
  • Xue, J.-h,Hall, P. Why Does Rebalancing Class-Unbalanced Data Improve AUC for Linear Discriminant Analysis?. 37. 1109-1112
  • Zitzler, E,Laumanns, M,Thiele, L. (2001). SPEA2: Improving the Strength Pareto Evolutionary Algorithm..