Describing Location Shifts with One Class Support Vector Machines

igescu, iulia (2020): Describing Location Shifts with One Class Support Vector Machines.

Abstract

The evolution of variables during location shifts (structural breaks) is of high interest to policy makers. I propose a novel approach to describe location shifts. I use two business surveys in the industry sector (faster soft indicators) to target the industrial production index (a slower hard indicator). Then I use One-Class Support Vector Machines on combinations of these two variables to identify if new observations act as ’novelties’ for the target variable, as observations coming from a different distribution. In that case, one would expect the onset/end of a location shift. Moreover, that gives insights into what role animal spirit, as manifested in survey data, plays in equilibrium formation (location shifts).

• Item Type: MPRA Paper
• Language: English
• Keywords: SVM, location shifts, novelites
• Subjects: C - Mathematical and Quantitative Methods > C2 - Single Equation Models ; Single Variables > C25 - Discrete Regression and Qualitative Choice Models ; Discrete Regressors ; Proportions ; Probabilities; C - Mathematical and Quantitative Methods > C5 - Econometric Modeling; C - Mathematical and Quantitative Methods > C5 - Econometric Modeling > C55 - Large Data Sets: Modeling and Analysis; E - Macroeconomics and Monetary Economics > E6 - Macroeconomic Policy, Macroeconomic Aspects of Public Finance, and General Outlook > E65 - Studies of Particular Policy Episodes

References

Castle, J., Clements, M., and Hendry, D. 2016. Overview of Forecasting Facing Breaks, Journal of Business Cycle, 12:3-23. Clements, M. and Hendry, D. 2001. Forecasting Non-stationary Economic Time Series. MIT Press. Farmer, R. and Woodford, M. 1997. Self-fulfilling Prophecies and the Business Cycle. Macroeconomic Dynamics I: 740-69. Farmer, R. 1999. Macroeconomics of Self-fulfilling Prophecies. MIT Press. Hendry, D. F. 1995. Dynamic Econometrics. Oxford: Oxford University Press. Rappoport, P. and Reichlin, L. 1989. Segmented Trends and Non-Stationary Time Series. Economic Journal 99: 168-177. Schoelkopf, B., Williamson, R., Smola, A., Shawe-Taylor, J., Platt, J. 2000. Support Vector Method for Novelty Detection. Advances in Neural Information Processing Systems: Proceedings of the 1999 Conference. http://papers.nips.cc/paper/1723-support-vector-method-fornovelty-detection.pdf. Smola, A. and Schoelkopf, B. 2004. A Tutorial on Support Vector Regression. Statistics and Computing 14: 199–222.