Fukushima: The Failure of Predictive Models

Brian Stacey

MPRA Paper from University Library of Munich, Germany

Abstract: Linear regression is an attempt to build a model from existing data. It uses the existing data to fit a linear equation that exhibits the least error from the actual points. Even within the existing data, a regression equation has limited ability to predict actual dependent variable values. Linear regression is limited in several ways; it assumes only linear relationships between variables, it is very sensitive to outliers, and data points must be independent. If any of these assumptions are not met the models get less accurate. Often random error (residuals) are enough to reduce the quality of fit of the equation to where it is unable to predict any values within the range of the original data except (x ̅,y ̅). Even with the best coefficient of determination (R2=1), data within the sample set may not be indicative of the condition outside the sample set.

Keywords: Regression; Residual Errors; Prediction; Earthquake; Fukushima (search for similar items in EconPapers)
JEL-codes: C13 C51 (search for similar items in EconPapers)
Date: 2015-04-01
References: View complete reference list from CitEc
Citations:

Downloads: (external link)
https://mpra.ub.uni-muenchen.de/69383/1/MPRA_paper_69383.pdf original version (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:pra:mprapa:69383

Access Statistics for this paper

More papers in MPRA Paper from University Library of Munich, Germany Ludwigstraße 33, D-80539 Munich, Germany. Contact information at EDIRC.
Bibliographic data for series maintained by Joachim Winter ().