Statistical Analysis Stack Exchange
Statistical Analysis Stack Exchange

Questions tagged [mars]

Multivariate adaptive regression splines (MARS) is a regression model that involves automatic term selection, including automatic selection of interactions. The name "MARS" is trademarked, so free-software implementations often use the term "Earth" instead.

26 questions
22
votes
1 answer

Can splines be used for prediction?

I cannot be specific about the nature of the data as it is proprietary, but suppose we have data like this: Each month, some people sign up for a service. Then, in each subsequent month, those people may upgrade the service, discontinue the service…
Peter Flom
  • 94,055
  • 35
  • 143
  • 276
6
votes
0 answers

State of the Art versions of Generalized Additive Models

Generalized Additive Models [Trevor Hastie and Robert Tibshirani 86] was well received with over 1335 Citations. I am also aware of the popular(?) version of GAM - the Multivariate Adaptive Regression Splines MARS by Friedman 91 - 4568 Citations. I…
Raghav RV
  • 111
  • 8
5
votes
1 answer

Are interactions with quadratic terms in MARS possible?

I am using the mars function of the R earth package. When the degree of interaction of MARS terms is at least 1, is it possible to have quadratic terms like x1*x1 in the model? I guess interactions between different hinge functions are possible.…
Patrick Balada
  • 609
  • 9
  • 16
5
votes
3 answers

eliminating outliers in MARS regression

I using the regression method called MARS, in R is it called earth and is located in the package earth, in order to find the best regression model for my datat. I know that this method is suitable for large data-sets, can handle NA and also decides…
Paul
  • 471
  • 1
  • 5
  • 19
3
votes
1 answer

Model averaging and intercept interpretation in Splines

I have two questions regarding Multivariate Adaptive Regression splines (MARS) 1) How is intercept interpreted in MARS? In linear regression, my intercept is shown in the image below In a linear regression, the intercept is simply the mean response…
user53020
  • 635
  • 1
  • 5
  • 15
3
votes
1 answer

Should I capitalize GAM, MARS and LOESS?

I have to mention GAM (generalized additive models), MARS (multivariate adaptive regression splines), and LOESS in an academic paper. I think GAM and MARS are capitalized in most cases, but not sure about LOESS. I saw LOESS, Loess, and loess all.
user67275
  • 871
  • 1
  • 8
  • 27
2
votes
1 answer

How well do Multivariate Adaptive Regression Splines work in high dimensional settings?

I have been reading the Hastie and Tibshirani book again lately, and I noticed in Chapter 9 that the mention the MARS algorithm: Multivariate Adaptive Regression Splines, which is a nonparametric method for fitting a curve to some data. My question…
krishnab
  • 1,008
  • 7
  • 12
2
votes
1 answer

MARS regression

I have a question about MARS, If I have for example 50 observations of 5 sensors with 5 signals and I tried to do regression with MARS. I found the model eliminate the 5th sensor readings as it is so near. So, the model is function of 4 sensor…
user225499
  • 21
  • 2
2
votes
0 answers

Interpretation of Multivariate Adaptive Regression Splines (MARS) with Multiple Predictors

When it comes to multiple predictors, I've seen conflicting interpretations of MARS models and hoping for some clarification. Some fake results, let's say it's predicting household income via years of age and education... estimate …
erebusgw
  • 21
  • 2
2
votes
0 answers

How are the values of knots in MARS calculated?

I'm wondering how the MARS algorithm decides on the specific values for the knots which are included in the hinge functions. Thank you very much.
Daniel
  • 35
  • 3
1
vote
1 answer

Predicting Probabilities of non-binary outcomes with multivariate adaptive regression splines (MARS) in R

I have three non-binary, discrete variables X0.1, X0.12 and X0.15. I want to model X0.1 as a MARS model of X0.12 and X0.15, then using the model answer queries of the form P(X0.1=x|X0.12=y, X0.15=z). This is what I have tried so far : > mars1 <-…
bayesianbot
  • 13
  • 3
1
vote
1 answer

Specifying no interaction in earth package in R

Sample data dat <- structure(list(location.id = structure(c(198L, 243L, 147L, 114L,152L, 117L, 151L, 129L, 148L, 154L, 233L, 12L, 226L, 129L, 146L,126L, 247L, 147L, 80L, 133L, 162L, 147L, 152L, 126L, 36L, 154L,147L, 166L, 132L, 241L, …
89_Simple
  • 751
  • 1
  • 9
  • 23
1
vote
1 answer

How does earth (multivariate adaptive regression splines) handle categorical predictors?

I wanted to understand more about whether my implementation of the multivariate adaptive regression splines is correct. I have crop yield data from multiple locations and year and I want to predict yield as a function of location, year and some…
89_Simple
  • 751
  • 1
  • 9
  • 23
1
vote
1 answer

Wrong functions fitted opposite to raw data

I need advice on fitted functions in the below image. I fitted the below function using a multivariate adaptive regression splines from the earth package in R. fit <- earth(x = pred, y = resp, keepxy = T,newvar.penalty = 2, degree= 1,pmethod =…
user53020
  • 635
  • 1
  • 5
  • 15
1
vote
1 answer

Multi-adaptive Regression Splines (MARS) equation interpretation

I was curious about certain aspects of multi-adaptive regression splines (MARS or earth in R). Taking the equation of MARS from the wiki page: ozone = 5.2 + + 0.93max(0,temp - 58) - 0.64max(0,temp - 68) -…
user53020
  • 635
  • 1
  • 5
  • 15
1
2