| Contents |
| In this work we applied several interpolation techniques including locally weighted scatterplot smoothing techniques (Lowess/Loess) to geophysical and high frequency financial data. The application of these methods to the two different data sets demonstrate that the overall methods are accurate and efficient. In addition, these methods are highly localized and data dependent so that results are dependent on the data trends. Unlike the previous modeling implementations, this modeling technique deals with the spatial analysis of the data although for the high frequency financial data we used the modified version of the non-parametric regression method to find out the curve of best fit. Overall this modeling approach proves out to be very reliable and useful for handling spatial data and time dependent financial data. |
|