Curve Fitting of Averaged Data
In manipulating data for estimating parameters for models of physical phenomena several methods of aggregation can be employed. This article shows the conditions under which three typical methods of aggregation of data for such purposes are equivalent. More importantly it shows when these methods are not equivalent. The methods investigated are: 1. Aggregating all the observations made by several individuals, and estimating the parameters. 2. Using the average of each individual's successive observation as data to make the parameter estimate. 3. Averaging the parameter estimates generated by each individual's data set. The authors point out that method one and method two are equivalent if the numbers of data points in each observation set are equal or if the fitted curve passes through the average observation of each individual data set. Method three is equivalent to the other two if the curve determined is a linear function of the parameter and the number of observations in each data set aggregated are equal.
- The estimation of complete aggregation structures
- scientific article; zbMATH DE number 3940577
- FITTING GENERATED AGGREGATION OPERATORS TO EMPIRICAL DATA
- Median aggregation of distribution functions
- The relative efficiencies of the aggregated generalized improved double parameter estimators of the parameters in the linear model with aggregated data
This page was built for publication: Curve Fitting of Averaged Data
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3324860)