摘要：In the research area of crop yield density estimation and in particular in risk analysis, little emphasis has been given to the appropriateness of transformation methods (e.g., removing a linear trend) and how such transformations impact the reliability of the empirical distribution functions and the resulting probability estimates. Similarly, there is little consensus on the impact of environmental variables (e.g., rainfall and temperature) on empirical distributions of yields. Using historical county corn yield data for Arkansas and Louisiana and nonparametric methods, this empirical analysis shed light on the importance of data transformation in crop risk analysis. Results demonstrate that inappropriate data treatment can lead to misestimation of probability density estimates.
关键词：Probability Density Estimation;Nonparametric;Kernel;Nonstationary;Unit Roots;Data Transformations;Corn Yields;Weather
