Question

In Nonlinear regression analysis, if we have a data set and if we need to model nonlinear, once we enter the data either in R or SPSS software we observe the results, but we cannot see the exact equation as we can write for linear regression.as an example if we want to see the relationship between Sales the response variable and the advertising budget for TV media and News paper we have to independent variables,once the data we entered how can we get a best nonlinear model for the data ,are we going to suggest for the model or what to do?

thanks

Answer 1

You can use a number of methods to check the quality of the model, you can use k fold cross validation to check the quality of fit of your model, you can use leave one out cross validation, all these are numerical measures to check the quality of your fit, which basically seperates the data into two parts, one is used to fit the data-the training dataset(it's generally consisted of the majority of the data) , and the other part of the data is used to check the quality of the fit -the test dataset(minority of the dataset).

For graphical measures,you can plot the fitted curve along with the prediction and confidence intervals , and for plotting for each degree of the equation you fit, which shows how good the curve actually captures the variability of the data,

All these can be done using available packages, and in case if you need some customisations for a specific dataset those can be easily done, at most writting your own programs which might take a bit of time but in the end is worthful,