Loss Functions for Detecting Outliers in Panel Data: An Introduction

Charles, Coleman (2003): Loss Functions for Detecting Outliers in Panel Data: An Introduction. Published in: The 13th Federal Forecasters Conference - 2003: Papers and Proceedings No. 2003 (2003): pp. 265-273.

Abstract

Loss functions are introduced for detecting outliers in panel data. The loss functions for nonnegative data take into account both the size of the base and the relative change. When the data generation processes take a particular form, an exact parametrization is available. The loss functions are extended to variables whose outlier criteria depend on another variable and to data of mixed sign. In the latter case, the geometry dictates one parametrization.

• Item Type: MPRA Paper
• Language: English
• Keywords: Estimates, forecasts, outliers, data quality, panel data
• Subjects: C - Mathematical and Quantitative Methods > C1 - Econometric and Statistical Methods and Methodology: General > C19 - Other; C - Mathematical and Quantitative Methods > C3 - Multiple or Simultaneous Equation Models ; Multiple Variables > C33 - Panel Data Models ; Spatio-temporal Models; C - Mathematical and Quantitative Methods > C5 - Econometric Modeling > C52 - Model Evaluation, Validation, and Selection

References

Barnett, Vic and Lewis, Toby (1994). Outliers in Statistical Data, 3rd edition, John Wiley & Sons, New York.

Coleman, Charles D., (2000). “Evaluating and Optimizing Population Projections Using Loss Functions,” Federal Forecasters Conference 2000: Papers and Proceedings, Washington: U.S Department of Education, Office of Educational Research and Improvement, 27-32.

Coleman, Charles D., (2002). “Optimizing Population Projections Using Loss Functions When the Base Populations are Subject to Revision,” Federal Forecasters Conference 2002: Papers and Proceedings, Washington: U.S Department of Education, Office of Educational Research and Improvement, 27-32.

Coleman, Charles D. (2003). “Loss Functions for Assessing the Accuracy of Cross-Sectional Predictions,” manuscript, U.S. Census Bureau.

Coleman, Charles D. and Bryan, Thomas (2003). “Loss Functions for Detecting Outliers in Panel Data when the Data May Change Sign,” manuscript, U.S. Census Bureau.

Coleman, Charles D., Bryan, Thomas and Devine, Jason (2003). “Loss Functions for Detecting Outliers in Panel Data,” manuscript, U.S. Census Bureau.

DuMouchel, William (1999). “Bayesian Data Mining in Large Frequency Tables, With an Application to the FDA Spontaneous Reporting System,” The American Statistician 53, 177-188.

Hoaglin, David C. (1983). “Letter Values: A Set of Selected Order Statistics.” In, Hoaglin, David C., Frederick Mosteller and John W. Tukey [eds.], Understanding Robust and Exploratory Data Analysis, Wiley, New York.

Lindley, D.V. (1953). “Statistical Inference,” Journal of the Royal Statistical Society, Series B 15, 30-76.

Rousseeuw, Peter J., Ruts, Ida and Tukey, John W. (1999). “The Bagplot: A Bivariate Boxplot,” The American Statistician 53, 382-387.

Rudas, T., Clogg, C. C. and Lindsay, B. G. (1994). “A New Index of Fit Based on Mixture Methods for the Analysis of Contingency Tables,” Journal of the Royal Statistical Society, Series B 56, 623-639.

Tukey, John W. (1977). Exploratory Data Analysis, Addison-Wesley, Reading, Massachussetts.