Nonlinear Fitting Lecture
Transcript of Nonlinear Fitting Lecture
The importance of using the nonlinear equation rather than the linearly transformed equation.
Visualizing how the algorithm arrives at the minimizedsum of squared error terms by trying different combinationsof the parameters (A and B in this case)
How should the data in a “residuals” plot appear?
The importance of providing reasonable initial “guesses”of the parameters of interest
The error estimates of the parameters are helpful, but theyshould be regarded as underestimates of the true error. TheSolver algorithm in Excel does not provide error estimates
An “outlier” has more influence in determining the best-fit line when there are relatively fewer data points
An “outlier” has more influence in determining the best-fit line when there are relatively fewer data points