Imperial College London

About: Imperial College London is a education organization based out in London, Westminster, United Kingdom. It is known for research contribution in the topics: Population & Medicine. The organization has 90019 authors who have published 209164 publications receiving 9337534 citations. The organization is also known as: Imperial College of Science, Technology and Medicine & Imperial College.

Principles of Data Mining

Abstract: The book consists of three sections. The first, foundations, provides a tutorial overview of the principles underlying data mining algorithms and their application. The presentation emphasizes intuition rather than rigor. The second section, data mining algorithms, shows how algorithms are constructed to solve specific problems in a principled manner. The algorithms covered include trees and rules for classification and regression, association rules, belief networks, classical statistical models, nonlinear models such as neural networks, and local "memory-based" models. The third section shows how all of the preceding analysis fits together when applied to real-world data mining problems. Topics include the role of metadata, how to handle missing data, and data preprocessing.

KAZE features

Abstract: In this paper, we introduce KAZE features, a novel multiscale 2D feature detection and description algorithm in nonlinear scale spaces. Previous approaches detect and describe features at different scale levels by building or approximating the Gaussian scale space of an image. However, Gaussian blurring does not respect the natural boundaries of objects and smoothes to the same degree both details and noise, reducing localization accuracy and distinctiveness. In contrast, we detect and describe 2D features in a nonlinear scale space by means of nonlinear diffusion filtering. In this way, we can make blurring locally adaptive to the image data, reducing noise but retaining object boundaries, obtaining superior localization accuracy and distinctiviness. The nonlinear scale space is built using efficient Additive Operator Splitting (AOS) techniques and variable conductance diffusion. We present an extensive evaluation on benchmark datasets and a practical matching application on deformable surfaces. Even though our features are somewhat more expensive to compute than SURF due to the construction of the nonlinear scale space, but comparable to SIFT, our results reveal a step forward in performance both in detection and description against previous state-of-the-art methods.

Pore-network extraction from micro-computerized-tomography images

Abstract: Network models that represent the void space of a rock by a lattice of pores connected by throats can predict relative permeability once the pore geometry and wettability are known. Micro-computerized-tomography scanning provides a three-dimensional image of the pore space. However, these images cannot be directly input into network models. In this paper a modified maximal ball algorithm, extending the work of Silin and Patzek [D. Silin and T. Patzek, Physica A 371, 336 (2006)], is developed to extract simplified networks of pores and throats with parametrized geometry and interconnectivity from images of the pore space. The parameters of the pore networks, such as coordination number, and pore and throat size distributions are computed and compared to benchmark data from networks extracted by other methods, experimental data, and direct computation of permeability and formation factor on the underlying images. Good agreement is reached in most cases allowing networks derived from a wide variety of rock types to be used for predictive modeling.