Discussion Of: a Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures over the Last 1000 Years
نویسندگان
چکیده
Professors McShane and Wyner have written a thought-provoking paper that intends to challenge some of the conventional wisdom in the paleoclimate literature. Rather than commenting on the merits of the entire methodology we focus on one topic. Namely, we discuss theoretical and practical aspects of the use of the least absolute shrinkage and selection operator [Tibshirani (1996)], more popularly known as the “Lasso,” in the context of paleoclimate reconstruction. It is important to acknowledge at first sight that the Lasso seems like a natural candidate in the paleoclimate context, since one is immediately faced with a larger number of proxies, compared to the number of data points [e.g., in McShane and Wyner (2010) (hereafter MW), Section 3.2, the response variable is of length 149 whereas there are 1138 predictors]. It is clear that standard regression-based variable selection techniques will not work. The sheer number of predictors does indeed warrant a need for regularization. Many techniques are available for such problems, including popular methods such as ridge regression and principal component regression. As pointed out by MW the “Lasso tends to choose sparse β̂ Lasso thus serving as a variable selection methodology and alleviating the p≫ n problem.” This point is very well taken. The model selection capability of the Lasso has made it very relevant in this era of high throughput data and rapidly changing information technology. Consequently the Lasso has been useful in biomedical and genomic applications where genes are often in the
منابع مشابه
Further Information For: a Comment on “a Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures over the Last 1000 Years Reliable?” by Mcshane and Wyner
متن کامل
DISCUSSION OF: A STATISTICAL ANALYSIS OF MULTIPLE TEMPERATURE PROXIES: ARE RECONSTRUCTIONS OF SURFACE TEMPERATURES OVER THE LAST 1000 YEARS RELIABLE?1 BY ALEXEY KAPLAN2 Lamont–Doherty Earth Observatory
McShane and Wyner (2011) (hereinafter MW2011) demonstrated that in many cases a comprehensive data set of p = 1138 proxies [Mann et al. (2008)] did not predict Northern Hemisphere (NH) mean temperatures significantly better than random numbers. This fact is not very surprising in itself: the unsupervised selection of good predictors from a set of p n proxies of varying sensitivities might be to...
متن کاملA Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures over the Last 1000 Years Reliable? by Blakeley
Predicting historic temperatures based on tree rings, ice cores, and other natural proxies is a difficult endeavor. The relationship between proxies and temperature is weak and the number of proxies is far larger than the number of target data points. Furthermore, the data contain complex spatial and temporal dependence structures which are not easily captured with simple models. In this paper,...
متن کاملA Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures over the Last 1000 Years Reliable?
Predicting historic temperatures based on tree rings, ice cores, and other natural ”proxies” is a difficult endeavor. The relationship between proxies and temperature is weak and the number of proxies is far larger than the number of target data points. Furthermore, the data contain complex spatial and temporal dependence structures which are not easily captured with simple models. In this talk...
متن کاملDiscussion Of: a Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures over the Last 1000 Years Reliable? by Gavin
McShane and Wyner (2011) (henceforth MW) analyze a dataset of “proxy” climate records previously used by Mann et al. (2008) (henceforth M08) to attempt to assess their utility in reconstructing past temperatures. MW introduce new methods in their analysis, which is welcome. However, the absence of both proper data quality control and appropriate “pseudoproxy” tests to assess the performance of ...
متن کاملDiscussion Of: a Statistical Analysis of Multiple Temperature Proxies: Are Reconstructions of Surface Temperatures
McShane and Wyner (2010; hereinafter MW2010) demonstrated that in many cases a comprehensive data set of p =1138 proxies (Mann et al., 2008) did not predict Northern Hemisphere (NH) mean temperatures significantly better than random numbers. This fact is not very surprising in itself: the unsupervised selection of good predictors from a set of p≫n proxies of varying sensitivities might be too c...
متن کامل