This paper presents a method of learning distributed representation for multi-field categorical data, which is a common data format with various applications such as recommender systems, social link prediction, and computational advertising. The success of non-linear models, e.g., factorisation machines, boosted trees, has proved the potential of exploring the interactions among inter-field cat...

An open-source software including an easy-to-use graphical user interface (GUI) has been developed for processing, modeling and mapping of gravity and magnetic data. The program, called Potensoft, is a set of functions written in MATLAB. The most common application of Potensoft is spatial and frequency domain filtering of gravity and magnetic data. The GUI helps the user easily change all the r...

Measures of the local phase of potential fields can be a useful aid to their interpretation. There are several variations in use, such as the tilt angle, tilt derivative and the Theta map. This paper compares the results of these filters, and introduces some new phase-based filters which show improved performance as edge detectors in different ways. The filters are demonstrated on synthetic gra...

The RKHS-based optimal image interpolation method, presented by Chen and de Figueiredo (1993), is applied to scattered potential field measurements. The RKHS which admits only interpolants consistent with Laplace's equation is defined and its kernel, derived. The algorithm is compared to bicubic spline interpolation, and is found to yield vastly superior results.

S U M M A R Y Edge effects are distortions at the edge of a domain which are artifacts of the implicit assumptions of a numerical algorithm or of the limited span of data used to solve the problem (Sheriff, 1991, p. 96). During wavenumber domain filtering of potential-field data, (e.g., reduction to the pole, pseudogravity, and directional derivative, etc.), edge effect distortions are usually ...

In providing a meta-analysis of a series of workshop papers and questions arising on the emergent field of learning analytics, this paper contributes to the ongoing formation of a shared research agenda. The first ICCE Learning Analytics workshop in 2014 demonstrated the effectiveness of a focused questioning session for collecting relevant data beyond the content of the papers themselves. In D...

Given a Poisson potential field (or its derivatives) on a plane in free space, the entire ambiguity domain (i.e., the set of all source distributions that could generate the field) is easily reconstructed in the Fourier domain. The resulting models show a range of variability far larger than normally expected. In essence, potential fields in free space only contain enough information to determi...

We introduce 2D potential field migration for the interpretation of gravity gradiometry data. This method is based on a direct integral transformation of the observed gravity gradients into subsurface density distributions which can be used for realtime imaging or as an a priori model for subsequent 3D regularized inversion. Potential field migration is very fast and stable with respect to nois...

This paper presents a statistical method for deriving optimal spatial sampling schemes. It focuses on ground verification of minerals derived from hyperspectral data. Spectral angle mapper (SAM) and spectral feature fitting (SFF) classification techniques were applied to obtain rule mineral images. The rule images provide weights that are utilized in simulated annealing. A HyMAP 126channel airb...