A Computational Intelligence Based Online Data Imputation Method: An Application For Banking
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
All the imputation techniques proposed so far in literature for data imputation are offline techniques as they require a number of iterations to learn the characteristics of data during training and they also consume a lot of computational time. Hence, these techniques are not suitable for applications that require the imputation to be performed on demand and near real-time. The paper proposes a computational intelligence based architecture for online data imputation and extended versions of an existing offline data imputation method as well. The proposed online imputation technique has 2 stages. In stage 1, Evolving Clustering Method (ECM) is used to replace the missing values with cluster centers, as part of the local learning strategy. Stage 2 refines the resultant approximate values using a General Regression Neural Network (GRNN) as part of the global approximation strategy. We also propose extended versions of an existing offline imputation technique. The offline imputation techniques employ K-Means or K-Medoids and Multi Layer Perceptron (MLP) or GRNN in Stage-1 and Stage-2 respectively. Several experiments were conducted on 8 benchmark datasets and 4 bank related datasets to assess the effectiveness of the proposed online and offline imputation techniques. In terms of Mean Absolute Percentage Error (MAPE), the results indicate that the difference between the proposed best offline imputation method viz., K-Medoids+GRNN and the proposed online imputation method viz., ECM+GRNN is statistically insignificant at a 1% level of significance. Consequently, the proposed online technique, being less expensive and faster, can be employed for imputation instead of the existing and proposed offline imputation techniques. This is the significant outcome of the study. Furthermore, GRNN in stage-2 uniformly reduced MAPE values in both offline and online imputation methods on all datasets.
Abstract
All the imputation techniques proposed so far in literature for data imputation are offline techniques as they require a number of iterations to learn the characteristics of data during training and they also consume a lot of computational time. Hence, these techniques are not suitable for applications that require the imputation to be performed on demand and near real-time. The paper proposes a computational intelligence based architecture for online data imputation and extended versions of an existing offline data imputation method as well. The proposed online imputation technique has 2 stages. In stage 1, Evolving Clustering Method (ECM) is used to replace the missing values with cluster centers, as part of the local learning strategy. Stage 2 refines the resultant approximate values using a General Regression Neural Network (GRNN) as part of the global approximation strategy. We also propose extended versions of an existing offline imputation technique. The offline imputation techniques employ K-Means or K-Medoids and Multi Layer Perceptron (MLP) or GRNN in Stage-1 and Stage-2 respectively. Several experiments were conducted on 8 benchmark datasets and 4 bank related datasets to assess the effectiveness of the proposed online and offline imputation techniques. In terms of Mean Absolute Percentage Error (MAPE), the results indicate that the difference between the proposed best offline imputation method viz., K-Medoids+GRNN and the proposed online imputation method viz., ECM+GRNN is statistically insignificant at a 1% level of significance. Consequently, the proposed online technique, being less expensive and faster, can be employed for imputation instead of the existing and proposed offline imputation techniques. This is the significant outcome of the study. Furthermore, GRNN in stage-2 uniformly reduced MAPE values in both offline and online imputation methods on all datasets.
참고문헌 (41)
- M. Abdella, "The use of Genetic Algorithms and Neural Networks to approximate missing data in database", Computational Cybernetics, ICCC 2005, IEEE 3rd International Conference, pp.207~212 (2005)
- N. Ankaiah, "A novel soft computing hybrid for data imputation", DMIN (2011)
- P. C. Austin, "Bayesian modeling of missing data in clinical research", Computa-tional Statics & Data Analysis, 49(3), pp.821~836 (2005)
- G. Batista, "A study of K-nearest neighbor as an imputation method, In Hybrid Intelligent Systems, Ser Front Artificial Intelligence Applications,vol. 87", pp.251~260 (2002)
- Experimental comparison of K-nearest neighbor and mean or mode imputation methods with the internal strategies used by C4.5 and CN2 to treat missing data, "G. Batista", University of Sao Paulo, ~ (2003)
- J. Chen, "A selective Bayes Classifier for classifying incomplete data based on gain ratio", Knowledge Based Systems, 21(7), pp.530~534 (2008)
- M. Cooke, "Handling missing data in speech recognition", International Conference on Spoken Language Process, pp.1555~1558 (1994)
- W. S. Desabro, "Missing data in product-concept testing", Decision Sciences, 17(2), pp.163~185 (1986)
- A.G. Di Nuovo, "Missing data analysis with fuzzy C-Means: A study of its application in a psychological scenario", Expert Systems With Applications, 38(6), pp.6793~6797 (2011)
- A.Elshorbagy, "Estimation of missing streamflow data using the prin-ciples of chaos theory", Journal of Hydrology, 255(1)통권 : 4, pp.123~133 (2002)
- B.Gabrys, "Neuro-Fuzzy approach to processing inputs with missing values in pattern recognition prob-lems", International Journal of Approximate Reasoning, 30(3), pp.149~179 (2002)
- P.J. Garcı´a-Laencina, "Pattern classification with miss-ing data: a review", Neural Computing & Applications, 19(2), pp.263~282 (2012)
- A.Gupta, "Estimating missing values using neural networks", Journal Of Operational Research Society, 47(2), pp.229~238 (1996)
- J. Han, "Data Mining concepts and techniques" (2008), Morgan Kufmann Publishers
- S. Henley, "The problem of missing data in geoscience databases", Computers & Geosciences, 32(9), pp.1368~1377 (2006)
- J. Jerez, "Missing data imputation in breast cancer progno-sis", BioMed'06 Proceedings of the 24th IASTED International Conference on Biomedical Engineering, pp.323~328 (2006)
- N. K. Kasabov, "DENFIS: Dynamic Evolving Neural-Fuzzy Inference System and Its Appli-cation for Time-Series Prediction", IEEE transactions on fuzzy systems, 10(2), pp.144~154 (2002)
- B. Kline, "Principles and Practice of Structural Equation Modelling", Guliford Press
- N. M. Laird, "Missing data in longitudinal studies", Statistics in Medicine, 7(1-2), pp.305~315 (1988)
- R. J. A. Little, "Statistical analysis with missing data", pp.2~250 (2002), Wiley
- H. LiuandS. Zhang, "Noisy data elimination using mutual k-nearest neighbor for classification mining", Journal of Systems and Software, 85(5), pp.1067~1074 (2012)
- J. B. MacQueen, "Some Methods for classification and Analysis of Multivariate Observa-tions", Proceedings of 5-th Berkeley Symposium on Mathematical Statistics and Probability, pp.281~297 (1967), University of California Press
- W. G. Madlow, "Incomplete data in sample surveys: Report and case Studies volume 1" (1983), Academic Press
- M. Marseguerra, "The autoassociative neural network in signal analysis. II. Application to on-line monitoring of a simulated BWR component", Annals of Nuclear Energy, 32(11), pp.1207~1223 (2002)
- T. Marwala, "Fault classification in structures with incomplete measured data using auto associative neural networks and genetic algorithm", Current Science India, 90(4), pp.542~548 (2006)
- P. Merlin, "X-SOM and L-SOM: A double classification ap-proach for missing value imputation", Neurocomputing, 73(7-9), pp.1103~1108 (2010)
- T. R. Nkuna, "Filling of missing rainfall data in Luvuvhu River Catchment using artificial neural networks", Physics and Chemistry of the Earth, Parts A, C, pp.2011, 36(14-15)[DOI]
- S. Nordbotten, "Neural network imputation applied to the Norwegian 1990 population census da-ta", Journal of Official Statistics, 12(4), pp.385~401 (1996)
- A.Ragel, "MVC—a preprocessing method to deal with missing values", Knowledge Based Systems, 12(5-6), pp.285~291 (1999)
- P. L. Roth, "Missing data in multiple item scales: a Monte Carlo analysis of missing data techniques", Organizational research methods, 2(3), pp.211~232 (1999)
- A.E. Rumlhart, "Learning internal representations by error propaga-tion", Parallel distributed processing: explorations in the microstructure of cognition, 1, pp.318~362 (1986)
- T. Samad, "Self-organization with partial data", Network: Computation in Neural Sys-tems, 3(2), pp.205~212 (1992)
- J.L. Schafer, "Analysisof incomplete multivariate data" (1997), Chapman & Hall
- P. K. Sharpe, "Dealing with missing values in neural network-based diagnostic sys-tems", Neural Computing & Applications, 3(2), pp.73~77 (1995)
- A.L. Silva-Ramírez, "Missing value im-putation on missing completely at random data using multilayer perceptrons", Neural Networks, 24(1), pp.121~129 (2011)
- Q. Song, "A new imputation method for small software project data sets", Journal of Systems and Software, 80(1), pp.1~62 (2007)
- D. F. Specht, "A General Regression Neural Network", IEEE transactions on neural networks, 2(6), pp.568~576 (1991)
- O. Troyanskaya, "Missing value estimation methods for DNA microarrays", Bioinformatics, 17(6), pp.520~525 (2001)
- I. Wasito, "Nearest Neighbor approach in the least-squares data imputation algorithms", Information Sciences, 16(1-2), pp.1~25 (2005)
- I. Wasito, "Nearest Neighbor in the least-squares data imputation algorithms with differ-ent missing patterns", Computational Statistics and Data Analysis, 50(4), pp.926~949 (2005)
- S. Y. Yoon, "Training algorithm with incomplete data for feed-forward neural net-works", Neural Processing Letters, 10(3), pp.171~179 (1999)