How would you like a modeling technique that provides all of the following:
- Offers the flexibility to build linear and nonlinear models for both regression and classification
- Can support variable interaction terms
- Is simple to understand and explain
- Requires little data processing
- Handles all types of data: numeric and categorical
- Performs well on unseen data, that is, it does well in a bias-variance trade-off
If that all sounds appealing, then I cannot recommend the use of MARS models enough. I've found them to perform exceptionally well. In fact, in a past classification problem of mine, they outperformed both a random forest and boosted trees on test/validation data.
To understand MARS is quite simple: ...
