Page 13 - Demo
P. 13

 Generalized Additive Models are able to reconstruct the source with high accuracy
β€’ Generalized Additive Models (GAMs) are less prone to overfitting than many ML algorithms
β€’ Widely applicable, from nonlinear multivariate regression problems to cluster analysis
β€’ Linear GAMS are generally of the form:
𝑔𝔼𝑦𝑋 =𝛽'+H𝑓(𝑋( (
β€’ 𝑔 is the β€œlink” function which relates b-spline functions of predictor variables (𝑓((𝑋()) to the expected value of the dependent variable
(𝔼 𝑦 𝑋 )
    Presenter: Gabriel Player
13
   11   12   13   14   15