When is the additive model appropriate

In this model, the trend and seasonal components are multiplied and then added to the error component. Choose the multiplicative model when the magnitude of the seasonal pattern in the data depends on the magnitude of the data. In other words, the magnitude of the seasonal pattern increases as the data values increase, and decreases as the data values decrease.

Choose the additive model when the magnitude of the seasonal pattern in the data does not depend on the magnitude of the data. Other options we could throw into this process, and usually do, would be to fit a polynomial regression for each neighborhood, and weight observations the closer they are to the value, with less weight given to more distant observations. The above plot shows the result from such a fitting process.

For comparison, the regular regression fit is also provided. The next figure regards a data set giving a series of measurements of head acceleration in a simulated motorcycle accident 5 Time is in milliseconds, acceleration in g. Here we have data that are probably not going to be captured with simple transformations of the predictors.

We could try higher orders, which would help, but in the end we will need something more flexible, and generalized additive models can help us in such cases. Of the exponential family. See the scatterplots in the car package for example. This process is experimental and the keywords may be updated as the learning algorithm improves. This is a preview of subscription content, log in to check access. Personalised recommendations. Cite chapter How to cite? Oxford Reference.

Publications Pages Publications Pages. Recently viewed 0 Save Search. Your current browser may not support copying via this button. Subscriber sign in You could not be signed in, please check and try again. Username Please enter your Username. Password Please enter your Password.