Browsing University of Alaska Fairbanks by Subject "Statistics"
Now showing items 1-4 of 4
Bayesian predictive process models for historical precipitation data of Alaska and southwestern CanadaIn this paper we apply hierarchical Bayesian predictive process models to historical precipitation data using the spBayes R package. Classical and hierarchical Bayesian techniques for spatial analysis and modeling require large matrix inversions and decompositions, which can take prohibitive amounts of time to run (n observations take time on the order of n3). Bayesian predictive process models have the same spatial framework as hierarchical Bayesian models but fit a subset of points (called knots) to the sample which allows for large scale dimension reduction and results in much smaller matrix inversions and faster computing times. These computationally less expensive models allow average desktop computers to analyze spatially related datasets in excess of 20,000 observations in an acceptable amount of time.
Extending the Lattice-Based Smoother using a generalized additive modelThe Lattice Based Smoother was introduced by McIntyre and Barry (2017) to estimate a surface defined over an irregularly-shaped region. In this paper we consider extending their method to allow for additional covariates and non-continuous responses. We describe our extension which utilizes the framework of generalized additive models. A simulation study shows that our method is comparable to the Soap film smoother of Wood et al. (2008), under a number of different conditions. Finally we illustrate the method's practical use by applying it to a real data set.
Martingales in mark-recapture experiments with constant recruitment and survivalThe method known as mark-recapture has been used for almost one hundred years in assessing animal populations. For many years, these models were restricted to closed populations; no changes to the population were assumed to occur through either migration or births and deaths. Numerous estimators for the closed population have been proposed through the years, some of the most recent by Paul Yip which make use of martingales to derive the necessary estimates. The independently derived Jolly-Seber model (1965) was the first to address the open population situation. That method as originally proposed is cumbersome mathematically due to the large number of parameters to be estimated as well as the inability to obtain estimates until the end of a series of capture events since some of the "observed" variables necessary are prospective. It also is cumbersome for the biologist in the field as individual marks and capture histories are required for each animal. Variations have been proposed through the years which hold survival and/or capture probabilities constant across capture occasions. Models based on log-linear estimators have also been proposed (Cormack 1989). This paper builds on the closed population work of Yip in using martingale-based conditional least squares to estimate population parameters for an open population where it is assumed recruitment of new individuals into the population is constant from one capture occasion to the next, and capture and survival probabilities are constant across capture occasions. It is an improvement over most other methods in that no detailed capture histories are needed; animals are simply noted as marked or unmarked. Performance of the estimator proposed is studied through computer simulation and comparison with classical estimators on actual data sets.
Participation, Preferences, and Characteristics of Outlying-Cabin Users in Alaska National ForestsThe development and management of public-use cabins have been planned, or at least considered, by several federal and state agencies in Alaska. This bulletin reports the results of a pilot study of the cabin program of the U.S. Forest Service. There are problems of aggregated data which did not allow for detailed analysis; however, the report does provide an overview of the Forest Service outlying cabin program-who uses it, how they use it, and how they feel about it. The manager should be careful in applying the results without consideration of the total recreational spectrum, i.e., where the cabin program fits within this spectrum, and its cost in terms of other recreation opportunities that may be specified. It is the opinion of the authors that it would be unwise to simply mass reproduce the outlying cabin program in all areas having periods of inclement weather. The study sampled only cabin users-not all users or potential users of the particular landscape setting. To over-emphasize an expanded cabin program would reduce the continuum of opportunities. While subsequent studies of the cabin user population would likely find this group to prefer the new program, the users who did not prefer it or who were unwilling to adopt to new conditions would have been displaced. Thus, while the results have some direct applicability, it is also important to consider the maintenance of the continuum of recreational opportunities, only one portion of which is covered by outlying cabins.