A fast dual algorithm for kernel logistic regression

This paper gives a new iterative algorithm for kernel logistic regression. It is based on the solution of a dual problem using ideas similar to those of the Sequential Minimal Optimization algorithm for Support Vector Machines. Asymptotic convergence of the algorithm is proved. Computational experiments show that the algorithm is robust and fast. The algorithmic ideas can also be used to give a fast dual algorithm for solving the optimization problem arising in the inner loop of Gaussian Process classifiers.

A model based factorization approach for dense 3D recovery from monocular video

Feature track matrix factorization based methods have been attractive solutions to the Structure-front-motion (Sfnl) problem. Group motion of the feature points is analyzed to get the 3D information. It is well known that the factorization formulations give rise to rank deficient system of equations. Even when enough constraints exist, the extracted models are sparse due the unavailability of pixel level tracks. Pixel level tracking of 3D surfaces is a difficult problem, particularly when the surface has very little texture as in a human face. Only sparsely located feature points can be tracked and tracking error arc inevitable along rotating lose texture surfaces. However, the 3D models of an object class lie in a subspace of the set of all possible 3D models. We propose a novel solution to the Structure-from-motion problem which utilizes the high-resolution 3D obtained from range scanner to compute a basis for this desired subspace. Adding subspace constraints during factorization also facilitates removal of tracking noise which causes distortions outside the subspace. We demonstrate the effectiveness of our formulation by extracting dense 3D structure of a human face and comparing it with a well known Structure-front-motion algorithm due to Brand.

RACK: RApid Clustering using K-means algorithm

The k-means algorithm is an extremely popular technique for clustering data. One of the major limitations of the k-means is that the time to cluster a given dataset D is linear in the number of clusters, k. In this paper, we employ height balanced trees to address this issue. Specifically, we make two major contributions, (a) we propose an algorithm, RACK (acronym for RApid Clustering using k-means), which takes time favorably comparable with the fastest known existing techniques, and (b) we prove an expected bound on the quality of clustering achieved using RACK. Our experimental results on large datasets strongly suggest that RACK is competitive with the k-means algorithm in terms of quality of clustering, while taking significantly less time.

The inside story: A survey of social work students' supervision and learning opportunities on placement

Practice learning accounts for half of the content of the bachelor of social work degree course requirements in Northern Ireland in their field education programmes and share a professional and ethical responsibility with practice teachers to provide appropriate learning environments to prepare students as competent and professional practitioners. The accreditation standards for practice learning require the placement to provide students with regular supervision and exposure to a range of learning strategies, but there is little research that actually identifies the types of placements offering this learning and the key activities provided. This paper builds on an Australian study and surveys social work students in two programmes in Northern Ireland about their exposure to a range of learning activities, how frequently they were provided and how it compares to what is required by the Northern Ireland practice standards. The results indicated that, although most students were satisfied with the supervision and support they received during their placement, the frequency of supervision and type of learning activities varied according to different settings, year levels and who provided the learning opportunities.

A real coded genetic algorithm for data partitioning and scheduling in networks with arbitrary processor release time

The problem of scheduling divisible loads in distributed computing systems, in presence of processor release time is considered. The objective is to find the optimal sequence of load distribution and the optimal load fractions assigned to each processor in the system such that the processing time of the entire processing load is a minimum. This is a difficult combinatorial optimization problem and hence genetic algorithms approach is presented for its solution.

Delay optimal control algorithm for a multiaccess fading channel with peak power constraint

We consider an optimal power and rate scheduling problem for a multiaccess fading wireless channel with the objective of minimising a weighted sum of mean packet transmission delay subject to a peak power constraint. The base station acts as a controller which, depending upon the buffer lengths and the channel state of each user, allocates transmission rate and power to individual users. We assume perfect channel state information at the transmitter and the receiver. We also assume a Markov model for the fading and packet arrival processes. The policy obtained represents a form of Indexability.

Pareto-optimal solutions for multi-objective optimization of fed-batch bioreactors using nondominated sorting genetic algorithm

Many optimal control problems are characterized by their multiple performance measures that are often noncommensurable and competing with each other. The presence of multiple objectives in a problem usually give rise to a set of optimal solutions, largely known as Pareto-optimal solutions. Evolutionary algorithms have been recognized to be well suited for multi-objective optimization because of their capability to evolve a set of nondominated solutions distributed along the Pareto front. This has led to the development of many evolutionary multi-objective optimization algorithms among which Nondominated Sorting Genetic Algorithm (NSGA and its enhanced version NSGA-II) has been found effective in solving a wide variety of problems. Recently, we reported a genetic algorithm based technique for solving dynamic single-objective optimization problems, with single as well as multiple control variables, that appear in fed-batch bioreactor applications. The purpose of this study is to extend this methodology for solution of multi-objective optimal control problems under the framework of NSGA-II. The applicability of the technique is illustrated by solving two optimal control problems, taken from literature, which have usually been solved by several methods as single-objective dynamic optimization problems. (C) 2004 Elsevier Ltd. All rights reserved.

Measurement of the top quark mass and pp̅ →tt̅ cross section in the all-hadronic mode with the CDF II detector

We present a measurement of the top quark mass and of the top-antitop pair production cross section using p-pbar data collected with the CDFII detector at the Tevatron Collider at the Fermi National Accelerator Laboratory and corresponding to an integrated luminosity of 2.9 fb-1. We select events with six or more jets satisfying a number of kinematical requirements imposed by means of a neural network algorithm. At least one of these jets must originate from a b quark, as identified by the reconstruction of a secondary vertex inside the jet. The mass measurement is based on a likelihood fit incorporating reconstructed mass distributions representative of signal and background, where the absolute jet energy scale (JES) is measured simultaneously with the top quark mass. The measurement yields a value of 174.8 +- 2.4(stat+JES) ^{+1.2}_{-1.0}(syst) GeV/c^2, where the uncertainty from the absolute jet energy scale is evaluated together with the statistical uncertainty. The procedure measures also the amount of signal from which we derive a cross section, sigma_{ttbar} = 7.2 +- 0.5(stat) +- 1.0 (syst) +- 0.4 (lum) pb, for the measured values of top quark mass and JES.

KAUPPAKESKUS OSANA KESTÄVÄÄ KAUPUNKIRAKENNETTA - Kauppakeskustoimijoiden näkemyksiä kauppakeskusten sijoittumisen kestävyydestä

The location and location guidance of shopping centers has been under much public discussion in Finland in the recent years. The Ministry of the Environment has expressed concern over the sustainability of ‘out-of-town’ shopping centers. Shopping centers outside the urban form are seen to cause more traffic, thus contributing to climate change by increasing carbon dioxide emissions. The sustainability of urban form has been researched in several studies and factors like urban density, public transport and a comfortable living environment were found to be the most important. This study presents the views of Finnish shopping center stakeholders on the sustainability of shopping center locations. These views were gathered using focus groups. Stakeholders included managers, consultants, investors, developers, architects and tenants of shopping centers and public sector actors dealing with shopping industry. As one theme in the discussions, participants were asked to present their views on the sustainability of shopping centers’ current locations. The study is part of the Aalto University of Technology KOKKKA project, which has its main focus upon shopping centers and sustainability. Shopping centers were seen to affect sustainability mainly through their location. A sustainable location was thought of as one that involved locating in an economically successful place, inside the urban form. A sustainable location was also easily accessible, with good access via public transport and the shopping center also had to create comfortable living environment in its surroundings. The views of the focus groups participants are similar to the views in sustainable urban structure theories and, inter alia, Finland’s national sustainable development strategy.

The effect of viscosity and surface tension on the Taylor instability of the surface of a cavitation bubble*

This paper is devoted to a consideration of the following problem: A spherical mass of fluid of density varrho1, viscosity μ1 and external radius R is surrounded by a fluid of density varrho2 and viscosity μ2.The fluids are immiscible and incompressible. The interface is accelerated radially by g1: to study the effect of viscosity and surface tension on the stability of the interface. By analyzing the problem in spherical harmonics the mathematical problem is reduced to one of solution of the characteristic determinant equation. The particular case of a cavity bubble, where the viscosity μ1 of the fluid inside the bubble is negligible in comparison with the viscosity μ2 of the fluid outside the bubble, is considered in some detail. It is shown that viscosity has a stabilizing role on the interface; and when g1 > T(n − 1) (n + 2)/R2(varrho2 − varrho1) the stabilizing role of both viscosity and surface tension is more pronounced than would result when either of them is taken individually.

Study of multi-muon events produced in p\bar{p} interactions at \sqrt{s}=1.96 TeV

We report the results of a study of multi-muon events produced at the Fermilab Tevatron collider and acquired with the CDF II detector using a dedicated dimuon trigger. The production cross section and kinematics of events in which both muon candidates are produced inside the beam pipe of radius 1.5 cm are successfully modeled by known processes which include heavy flavor production. In contrast, we are presently unable to fully account for the number and properties of the remaining events, in which at least one muon candidate is produced outside of the beam pipe, in terms of the same understanding of the CDF II detector, trigger, and event reconstruction.

A Geometric Algorithm for Learning Oblique Decision Trees

In this paper we present a novel algorithm for learning oblique decision trees. Most of the current decision tree algorithms rely on impurity measures to assess goodness of hyperplanes at each node. These impurity measures do not properly capture the geometric structures in the data. Motivated by this, our algorithm uses a strategy, based on some recent variants of SVM, to assess the hyperplanes in such a way that the geometric structure in the data is taken into account. We show through empirical studies that our method is effective.

Single-File Diffusion of Water Inside Narrow Carbon Nanorings

We use atomistic molecular dynamics (MD) simulations to study the diffusion of water molecules confined inside narrow (6,6) carbon nanorings. The water molecules form two oppositely polarized chains. It is shown that the effective interaction between these two chains is repulsive in nature. The computed mean-squared displacement (MSD) clearly shows a scaling with time similar to t(1/2), which is consistent with single-file diffusion (SFD). The time up to which the water molecules undergo SFD is shown to be the lifetime of the water molecules inside these chains. Simulations of "uncharged" water molecules inside the nanoring show the formation of several water chains and yield SFD. These observations conclusively prove that the diffusion is Fickian when there is a single chain of water and SFD is observed only when two or more chains are present.

Mechanism of glycine uptake by the silk glands of the silkworm Bombyx mori L.

The transport of glycine in vitro into the silk glands of the silkworm has been studied. Glycine accumulates inside the tissue to a concentration higher than that present outside, indicating an active transport mechanism. The kinetics of uptake show a biphasic curve and two apparent Km values for accumulation, 0.33 mM and 5.00 mM. The effect of inhibitors on the energy metabolism of glycine transport is inconclusive. Exchange studies indicate the existence of two pools inside the gland, one that is easily removed by exchange and osmotic shock, and the other which is not. The results obtained conform with the carrier model of Britten and McClure concerning the amino-acid pool in E. coli.