A variation of the reduction-based approach to spatiotemporal interpolation (STI), in which time is treated independently from the spatial dimensions, is proposed in this paper. We reviewed and compared three widely-used spatial interpolation techniques: ordinary kriging, inverse distance weighting and the triangular irregular network. We also proposed a new distribution-based distance weightin...

[1] Inverse weighted distance and regression nonexact techniques were evaluated for interpolating methods snow water equivalent (SWE) across the entire Colorado River Basin of the western United States. A 1-km spacing was used for the gridding of snow telemetry (SNOTEL) measurements for the years 1993, 1998, and 1999, which on average, represented higher than average, average, and lower than av...

Since the introduction of LLE (Roweis and Saul, 2000) and Isomap (Tenenbaum et al., 2000), a large number of non-linear dimensionality reduction techniques (manifold learners) have been proposed. Many of these non-linear techniques can be viewed as instantiations of Kernel PCA; they employ a cleverly designed kernel matrix that preserves local data structure in the “feature space” (Bengio et al...

Let $$G = K_{n_1,n_2,\cdots ,n_t}$$ be a complete t-partite graph on $$n=\sum _{i=1}^t n_i$$ vertices. The distance between vertices i and j in G, denoted by $$d_{ij}$$ is defined to the length of shortest path j. squared matrix $$\Delta (G)$$ G $$n\times n$$ with $$(i,j)^{th}$$ entry equal 0 if $$i j$$ $$d_{ij}^2$$ \ne . We define energy $$E_{\Delta }(G)$$ sum absolute values its eigenvalues. ...

In one spatial dimension, a class of integrable manybody systems are the Calogero-Sutherland-Moser systems. They consist of many identical non-relativistic particles interacting through two-body potentials of the inverse square type and various extensions such as the inverse sine squared, the inverse hyperbolic squared, and the inverse squared Jacobian elliptic function [1 – 3]. The significanc...

Some of the simplest loss functions considered in Machine Learning are the square loss, the logistic loss and the hinge loss. The most common family of algorithms, including Gradient Descent (GD) with and without Weight Decay, always predict with a linear combination of the past instances. We give a random construction for sets of examples where the target linear weight vector is trivial to lea...

Coal exploration and mining in extensively drilled and sampled coal seams can benefit from 3D statistical facies interpolation. Starting from closely spaced core descriptions, and using interpolation methods, a 3D optimum and robust facies distribution model was obtained for a thick, heterogeneous coal seam zone deposited in the non-marine As Pontes basin (Oligocene-Early Miocene, NW Spain). Se...

Objectives. We investigated immune phenotypes of HIV+ patients who present late, considering late presenters (LPs, CD4+ < 350/μL and/or AIDS), advanced HIV disease (AHD, CD4+ < 200/μL and/or AIDS), and AIDS presenters (AIDS-defining condition at presentation, independently from CD4+). Methods. Patients newly diagnosed with HIV at our clinic between 2007-2011 were enrolled. Mann-Whitney/Chi-squa...