The long-term goal of this research is to construct a unified global and mesoscale nonhydrostatic numerical weather prediction (NWP) models for the U.S. Navy using new numerical methods specifically designed for modern computer architectures; this unified model is called the Nonhydrostatic Unified Model of the Atmosphere or NUMA. To take full advantage of distributedmemory computers, the global...

Highways and in particular their pavements are the fundamental components of the road network. They require continuous maintenance since they deteriorate due to changing traffic and environmental conditions. Monitoring methods and efficient pavement management systems are needed for optimizing maintenance operations. Pavement performance prediction models are useful tools for determining the op...

Intraseasonal and seasonal prediction provides important information for decision-making and resource management, and has received increased attention in recent years. Despite substantial progress in numerical modeling in the past few decades, skillful seasonal prediction remains a challenge for many models. Verification and evaluation of model forecasts can offer users necessary information on...

Modern retailers frequently use recommendation systems to suggest products of interest to a collection of consumers. A closely related task is ratings prediction, in which the system predicts a numerical rating that a user u will assign to a product p. In this paper, we build three ratings prediction models for a dataset of products and users from Amazon.com and Yelp.com. We evaluate the streng...

This work unit is part of a coupled 6.2-6.4 development effort to implement rapid transition of aerosol prediction and EO tactical decision aid capability to the Fleet and is carried out by NRL-Montery and SPAWAR-SSC. The coupled program will improve the components of the Target Acquisition Weather Software (TAWS) and couple operational EO tactical decision aids (TDA's) to numerical weather pre...

background : in medical studies, when the joint prediction about occurrence of two events should be anticipated, a statistical bivariate model is used. due to the limitations of usual statistical models, other methods such as artificial neural network (ann) and hybrid models could be used. in this paper, we propose a hybrid artificial neural network-genetic algorithm (ann-ga) model to predictio...

This study deals with the missing answers prediction problem. We address this problem using coupled analysis of ImageCLEF2014 dataset by representing it as a heterogeneous data, i.e., dataset in the form of matrices. We propose to use an approach based on probabilistic interpretation of tensor factorization models, i.e., Generalized Coupled Tensor Factorization, which can simultaneously fit a l...