Mdl Knot Selection for Penalized Splines

نویسندگان

  • Antti Liski
  • Erkki P. Liski
چکیده

There exists a well known connection between penalized splines and mixed models. This connection makes it possible to exploit certain results derived for mixed models in the estimation of penalized splines. We have derived the Minimum Description Length (MDL) model selection criterion [1] for mixed models (see eg. [2]). In this paper we investigate the performance of the MDL criterion in fitting penalized splines. This is done through simulations. We focus on the problem of finding the optimal number of knots when using a truncated power basis. We compare the results of spline fitting by using the MDL criterion to results from other existing criteria.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Thin plate regression splines

I discuss the production of low rank smoothers for d ≥ 1 dimensional data, which can be fitted by regression or penalized regression methods. The smoothers are constructed by a simple transformation and truncation of the basis that arises from the solution of the thinplate spline smoothing problem, and are optimal in the sense that the truncation is designed to result in the minimum possible pe...

متن کامل

On The Asymptotics Of Penalized Splines

The asymptotic behaviour of penalized spline estimators is studied in the univariate case. We use B -splines and a penalty is placed on mth-order differences of the coefficients. The number of knots is assumed to converge to infinity as the sample size increases. We show that penalized splines behave similarly to Nadaraya-Watson kernel estimators with ‘equivalent’ kernels depending upon m. The ...

متن کامل

Data - driven Selection of the Spline Dimension in Penalized Spline Regression : Supplementary Material

where μ(.) is an unknown smooth function, the i are independent and identically distributed errors, i.e. i ∼ (0, σ2 ), and the xi take values in [0, 1], for simplicity. We estimate μ(.) by some high dimensional spline of the form X(x)β + ZK(x)uK . Here, X(.) is a low dimensional basis while ZK(.) is built from truncated polynomials, i.e. we set X(x) = (1, x, x2/2!, . . . , xq/q!) and ZK(x) = ( ...

متن کامل

Splines, Knots, and Penalties

Penalized splines have gained much popularity as a flexible tool for smoothing and semi-parametric models. Two approaches have been advocated: 1) use a B-spline basis, equally-spaced knots and difference penalties (Eilers and Marx, 1996) and 2) use truncated power functions, knots based on quantiles of the independent variable and a ridge penalty (Ruppert, Wand and Carroll, 2003). We compare th...

متن کامل

On the asymptotics of penalized spline smoothing

Abstract: This paper performs an asymptotic analysis of penalized spline estimators. We compare P -splines and splines with a penalty of the type used with smoothing splines. The asymptotic rates of the supremum norm of the difference between these two estimators over compact subsets of the interior and over the entire interval are established. It is shown that a Pspline and a smoothing spline ...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2008