An effective semi-supervised clustering framework integrating pairwise constraints and attribute preferences

Both the instance level knowledge and the attribute level knowledge can improve clustering quality, but how to effectively utilize both of them is an essential problem to solve. This paper proposes a wrapper framework for semi-supervised clustering, which aims to gracely integrate both kinds of priori knowledge in the clustering process, the instance level knowledge in the form of pairwise constraints and the attribute level knowledge in the form of attribute order preferences. The wrapped algorithm is then designed as a semi-supervised clustering process which transforms this clustering problem into an optimization problem. The experimental results demonstrate the effectiveness and potential of proposed method.

Dirichlet process mixture models with pairwise constraints for data clustering

The Dirichlet process mixture (DPM) model, a typical Bayesian nonparametric model, can infer the number of clusters automatically, and thus performing priority in data clustering. This paper investigates the influence of pairwise constraints in the DPM model. The pairwise constraint, known as two types: must-link (ML) and cannot-link (CL) constraints, indicates the relationship between two data points. We have proposed two relevant models which incorporate pairwise constraints: the constrained DPM (C-DPM) and the constrained DPM with selected constraints (SC-DPM). In C-DPM, the concept of chunklet is introduced. ML constraints are compiled into chunklets and CL constraints exist between chunklets. We derive the Gibbs sampling of the C-DPM based on chunklets. We further propose a principled approach to select the most useful constraints, which will be incorporated into the SC-DPM. We evaluate the proposed models based on three real datasets: 20 Newsgroups dataset, NUS-WIDE image dataset and Facebook comments datasets we collected by ourselves. Our SC-DPM performs priority in data clustering. In addition, our SC-DPM can be potentially used for short-text clustering.

Integrating instance-level and attribute-level knowledge into document clustering

In this paper, we present a document clustering framework incorporating instance-level knowledge in the form of pairwise constraints and attribute-level knowledge in the form of keyphrases. Firstly, we initialize weights based on metric learning with pairwise constraints, then simultaneously learn two kinds of knowledge by combining the distance-based and the constraint-based approaches, finally evaluate and select clustering result based on the degree of users’ satisfaction. The experimental results demonstrate the effectiveness and potential of the proposed method.

Groupwise registration of aerial images

This paper addresses the task of time separated aerial image registration. The ability to solve this problem accurately and reliably is important for a variety of subsequent image understanding applications. The principal challenge lies in the extent and nature of transient appearance variation that a land area can undergo, such as that caused by the change in illumination conditions, seasonal variations, or the occlusion by non-persistent objects (people, cars). Our work introduces several novelties: (i) unlike all previous work on aerial image registration, we approach the problem using a set-based paradigm; (ii) we show how local, pairwise constraints can be used to enforce a globally good registration using a constraints graph structure; (iii) we show how a simple holistic representation derived from raw aerial images can be used as a basic building block of the constraints graph in a manner which achieves both high registration accuracy and speed. We demonstrate: (i) that the proposed method outperforms the state-of-the-art for pair-wise registration already, achieving greater accuracy and reliability, while at the same time reducing the computational cost of the task; and (ii) that the increase in the number of available images in a set consistently reduces the average registration error.

Tensions and constraints for nurses in hospital-in-home programmes

Economic necessity constrains health-care expenditure and waiting lists for hospital treatments remain high. As a result, more care is delivered via alternative means, such as same-day surgery initiatives and home-care programmes. Acute care delivered in the home to patients who would otherwise require hospitalization is becoming an increasingly acceptable means of treatment. These Hospital-in-the-Home programmes offer increased comfort while delivering comparable outcomes to many patient groups. The purpose of this paper is to generate discussion concerning the tensions that exist for nurses who practice in the home under the auspices of acute-care institutions. Data drawn from field work that formed part of a critical ethnography is used to generate the discussion. The larger research project explored the constructions of the role of the nurse in four Hospital-in-the-Home programmes in Victoria, Australia. It will be argued that there is significant pressure exerted upon nurses to support the imperative to reduce bed days in acute hospitals by transferring people to their home. At times, this agenda clashes with the nurses’ professional commitment to provide holistic patient care yet the dilemmas are largely unacknowledged and/or unrecognized by the nurses despite the tension they generate.

Innovation chains: possibilities and constraints for critical perspectives on computers, difference and educational innovation

While declarations of ‘innovativeness’ are easily found in most educational contexts, it is significantly more difficult to locate detailed definitions of what educational innovation actually means. In this paper we are interested in identifying the extent to which mainstream takes on ‘innovation’ (as played out in contemporary technology and equity debates) reflect or respond to what we will define as the more innovative dimensions of innovation literature itself. Our aim throughout this paper, then, is to begin the complex process of developing a means for distinguishing between projects that are ‘badged’ as innovative and projects that are more demonstrably (and sustainably) innovative. In this process we will distinguish between what Shiv Visvanathn describes as “innovation chains”— dynamic, rhizomatic, transformative responses to the contemporary world that lead to fundamentally new ways of conceptualising technology, culture and difference—and the constraints—or chains—provided by dominant understandings of innovation: chains which anchor us to existing, hegemonic and limiting understandings of student
diversity and educational technology.

Polyhedral studies for minimum-span graph labelling with integer distance constraints

This paper studies the polytope of the minimum-span graph labelling problems with integer distance constraints (DC-MSGL). We first introduce a few classes of new valid inequalities for the DC-MSGL defined on general graphs and briefly discuss the separation problems of some of these inequalities. These are the initial steps of a branch-and-cut algorithm for solving the DC-MSGL. Following that, we present our polyhedral results on the dimension of the DC-MSGL polytope, and that some of the inequalities are facet defining, under reasonable conditions, for the polytope of the DC-MSGL on triangular graphs.

Passive angle measurement based localization consistency via geometric constraints

In this paper, we examine the geometric relations between various measured parameters and their corresponding errors in angle-measurement based emitter localization scenarios. We derive a geometric constraint formulating the relationship among the measurement errors in such a scenario. Using this constraint, we formulate the localization task as a constrained optimization problem that can be performed on the measurements in order to provide the optimal values such that the solution is consistent with the underlying geometry.

Late Paleozoic depositional history of the Tarim basin, northwest China: an integration of biostratigraphic and lithostratigraphic constraints

This study provides the first detailed lithostratigraphic and biostratigraphic constraints for improving stratigraphic resolution for hydrocarbon prospecting and exploration in the Tarim basin. A total of 49 stratigraphic units (38 formations and 11 members), ranging in age from the latest Devonian to Permian, are reviewed or redefined in terms of nomenclatures, lithology, age constraints, and lateral distributions based on the detailed field works or newly published data. Of these, the Piqiang Formation (new formation) is proposed to include the reefal carbonates of Asselian-Sakmarian age from the northern Tarim. The subsurface upper Paleozoic stratigraphic framework of the desert areas of the basin is also established for the first time. The high-resolution, basinwide stratigraphic correlations reveal that the sedimentation of the basin in the late Paleozoic was extremely uneven. Of these, the Famennian to Changhsingian successions are completely recorded in the south-western margin areas of the basin. Here, five eustatic sedimentary cycles are well recognizable, suggesting the sedimentation was more eustatically controlled and little affected by local tectonism. The late Paleozoic successions of both Kalpin and Taklimakan regions are commonly interrupted by major hiatuses at various horizons, suggesting that the sedimentation was apparently modified by local tectonism. Of these, the northward movement of the Tarim block and its subsequent collision with the Yili microcontinent (part of the Kazakhstan plate) may be principally accountable for the discrepancy in the sedimentation of the various regions in the basin in the late Paleozoic.

Socio-demographic constraints to travel behavior

Purpose – This study aims to ascertain the effect of socio-demographic constraints on dimension of travel choice. This study also seeks to derive personal ecological explanations for variation in travel preference, travel intention and travel choice behavior of a wide range of destinations.

Design/methodology/approach – A large representative sample of 49,105 Australian respondents is utilized. Binary logistic regression is used to determine the impact of constraint variables.

Findings – Age, income and life stage have significant differential and interactive effects on travel behavior. Socio-demographic variables act in different ways to constrain/free different types of travel behavior. However there are significant levels of travel by even the most constrained groups as well as significant amounts of non-travel by the least constrained sectors of our society. These impacts are country specific.

Research limitations/implications – The travel motivations of constraint groups need to be considered to order better understand travel behavior. Investigation of psychological and ecological facilitators and constraints to travel is needed.

Practical implications – This information is most useful for market segmentation and the development of constraint group destination marketing plans. Managers can use utilize such results to minimize the barriers to travel by particular groups.

Originality/value – This paper utilizes a large database to provide insights into the personal ecological constraints to travel.

Optimal range-difference-based localization considering geometrical constraints

This paper proposes a new type of algorithm aimed at finding the traditional maximum-likelihood (TML) estimate of the position of a target given time-difference-of-arrival (TDOA) information, contaminated by noise. The novelty lies in the fact that a performance index, akin to but not identical with that in maximum likelihood (ML), is a minimized subject to a number of constraints, which flow from geometric constraints inherent in the underlying problem. The minimization is in a higher dimensional space than for TML, and has the advantage that the algorithm can be very straightforwardly and systematically initialized. Simulation evidence shows that failure to converge to a solution of the localization problem near the true value is less likely to occur with this new algorithm than with TML. This makes it attractive to use in adverse geometric situations.

The effects of financial constraints on corporate policies in Australia

We jointly study the impact of financial constraints on Australian companies' investment decisions and demand for liquidity. By examining a large sample of Australian firms over the period 1990–2003, we find that financial constraints not only reduce the sensitivity of investment to the availability of internal funds, but also increase the responsiveness of cash holdings to internally generated cash flows. Further analysis shows that the impact of financial constraints varies across different cash flow states; that is, financial constraints have a small effect on corporate investment and cash policies when cash flows are positive. In contrast, the severity of constraints is high in negative cash flow years in which the cost disadvantage of external finance coincides with deteriorating operating performance.