Reconstruction of Cross-Sectional Missing Data Using Neural Networks

Citation

Gheyas IA & Smith L (2009) Reconstruction of Cross-Sectional Missing Data Using Neural Networks. In: Palmer-Brown D, Draganova C, Pimenidis E & Mouratidis H (eds.) Engineering Applications of Neural Networks: 11th International Conference, EANN 2009, London, UK, August 27-29, 2009. Proceedings. Communications in Computer and Information Science, 43. 11th International Conference, EANN 2009, London, UK, 27.08.2009-29.08.2009. Berlin Heidelberg: Springer, pp. 28-34. http://link.springer.com/chapter/10.1007/978-3-642-03969-0_3; https://doi.org/10.1007/978-3-642-03969-0_3

Abstract
The treatment of incomplete data is an important step in the pre-processing of data. We propose a non-parametric multiple imputation algorithm (GMI) for the reconstruction of missing data, based on Generalized Regression Neural Networks (GRNN). We compare GMI with popular missing data imputation algorithms: EM (Expectation Maximization) MI (Multiple Imputation), MCMC (Markov Chain Monte Carlo) MI, and hot deck MI. A separate GRNN classifier is trained and tested on the dataset imputed with each imputation algorithm. The imputation algorithms are evaluated based on the accuracy of the GRNN classifier after the imputation process. We show the effectiveness of our proposed algorithm on twenty-six real datasets.

Keywords
Missing values; imputation; multiple imputation; Generalized Regression Neural Networks

People

Professor Leslie Smith

Emeritus Professor, Computing Science