PDM 2006: Proceedings of the International Workshop on Data Mining

Recently major processor manufacturers have announced a dramatic shift in their paradigm to increase computing power over the coming years. Instead of focusing on faster clock speeds and more powerful single core CPUs, the trend clearly goes towards multi core systems. This will also result in a paradigm shift for the development of algorithms for computationally expensive tasks, such as data mining applications. Obviously, work on parallel algorithms is not new per se but concentrated efforts in the many application domains are still missing. Multi-core systems, but also clusters of workstations and even large-scale distributed computing infrastructures provide new opportunities and pose new challenges for the design of parallel and distributed algorithms. Since data mining and machine learning systems rely on high performance computing systems, research on the corresponding algorithms must be on the forefront of parallel algorithm research in order to keep pushing data mining and machine learning applications to be more powerful and, especially for the former, interactive. To bring together researchers and practitioners working in this exciting field, a workshop on parallel data mining was organized as part of PKDD/ECML 2006 (Berlin, Germany). The six contributions selected for the program describe various aspects of data mining and machine learning approaches featuring low to high degrees of parallelism: The first contribution focuses the classic problem of distributed association rule mining and focuses on communication efficiency to improve the state of the art. After this a parallelization technique for speeding up decision tree construction by means of thread-level parallelism for shared memory systems is presented. The next paper discusses the design of a parallel approach for dis- tributed memory systems of the frequent subgraphs mining problem. This approach is based on a hierarchical communication topology to solve issues related to multi-domain computational envi- ronments. The forth paper describes the combined use and the customization of software packages to facilitate a top down parallelism in the tuning of Support Vector Machines (SVM) and the next contribution presents an interesting idea concerning parallel training of Conditional Random Fields (CRFs) and motivates their use in labeling sequential data. The last contribution finally focuses on very efficient feature selection. It describes a parallel algorithm for feature selection from random subsets. Selecting the papers included in this volume would not have been possible without the help of an international Program Committee that has provided detailed reviews for each paper. We would like to also thank Matthew Otey who helped with publicity for the workshop.

Asset allocation, cross-class correlation and the structure of property returns

Practical applications of portfolio optimisation tend to proceed on a “top down” basis where funds are allocated first at asset class level (between, say, bonds, cash, equities and real estate) and then, progressively, at sub-class level (within property to sectors, office, retail, industrial for example). While there are organisational benefits from such an approach, it can potentially lead to sub-optimal allocations when compared to a “global” or “side-by-side” optimisation. This will occur where there are correlations between sub-classes across the asset divide that are masked in aggregation – between, for instance, City offices and the performance of financial services stocks. This paper explores such sub-class linkages using UK monthly stock and property data. Exploratory analysis using clustering procedures and factor analysis suggests that property performance and equity performance are distinctive: there is little persuasive evidence of contemporaneous or lagged sub-class linkages. Formal tests of the equivalence of optimised portfolios using top-down and global approaches failed to demonstrate significant differences, whether or not allocations were constrained. While the results may be a function of measurement of market returns, it is those returns that are used to assess fund performance. Accordingly, the treatment of real estate as a distinct asset class with diversification potential seems justified.

Duality, observability, and controllability for linear time-varying descriptor systems

A characterization of observability for linear time-varying descriptor systemsE(t)x(t)+F(t)x(t)=B(t)u(t), y(t)=C(t)x(t) was recently developed. NeitherE norC were required to have constant rank. This paper defines a dual system, and a type of controllability so that observability of the original system is equivalent to controllability of the dual system. Criteria for observability and controllability are given in terms of arrays of derivatives of the original coefficients. In addition, the duality results of this paper lead to an improvement on a previous fundamental structure result for solvable systems of the formE(t)x(t)+F(t)x(t)=f(tt).

The state estimation of flow demands in gas networks from sparse pressure telemetry

This paper presents novel observer-based techniques for the estimation of flow demands in gas networks, from sparse pressure telemetry. A completely observable model is explored, constructed by incorporating difference equations that assume the flow demands are steady. Since the flow demands usually vary slowly with time, this is a reasonable approximation. Two techniques for constructing robust observers are employed: robust eigenstructure assignment and singular value assignment. These techniques help to reduce the effects of the system approximation. Modelling error may be further reduced by making use of known profiles for the flow demands. The theory is extended to deal successfully with the problem of measurement bias. The pressure measurements available are subject to constant biases which degrade the flow demand estimates, and such biases need to be estimated. This is achieved by constructing a further model variation that incorporates the biases into an augmented state vector, but now includes information about the flow demand profiles in a new form.

The domestication of water: water management in the ancient world and its prehistoric origins in the Jordan Valley

The ancient civilizations were dependent upon sophisticated systems of water management. The hydraulic engineering works found in ancient Angkor (ninth to thirteenth century AD), the Aztec city of Tenochtitlan (thirteenth to fifteenth century AD), Byzantine Constantinople (fourth to sixth century AD) and Nabatean Petra (sixth century BC to AD 106) are particularly striking because each of these is in localities of the world that are once again facing a water crisis. Without water management, such ancient cities would never have emerged, nor would the urban communities and towns from which they developed. Indeed, the ‘domestication’ of water marked a key turning point in the cultural trajectory of each region of the world where state societies developed. This is illustrated by examining the prehistory of water management in the Jordan Valley, identifying the later Neolithic (approx. 8300–6500 years ago) as a key period when significant investment in water management occurred, laying the foundation for the development of the first urban communities of the Early Bronze Age.

Moderate drought alters biomass and depth distribution of fine roots in Norway spruce

A rain shelter experiment was conducted in a 90-year-old Norway spruce stand, in the Kysucké Beskydy Mts (Slovakia). Three rain shelters were constructed in the stand to prevent the rainfall from reaching the soil and to reduce water availability in the rhizosphere. Fine root biomass and necromass were repeatedly measured throughout a growing season by soil coring. We established the quantities of fine root biomass (live) and necromass (dead) at soil depths of 0-5, 5-15, 15-25, and 25-35 cm. Significant differences in soil moisture contents between control and drought plots were found in the top 15 cm of soil after 20 weeks of rainfall manipulation (lasting from early June to late October). Our observations show that even relatively light drought decreased total fine root biomass from 272.0 to 242.8 g m-2 and increased the amount of necromass from 79.2 to 101.2 g m-2 in the top 35 cm of soil. Very fine roots, i.e. those with diameter up to 1 mm, were more affected than total fine roots defined as 0-2 mm. The effect of reduced water availability was depth-specific, as a result we observed a modification of vertical distribution of fine roots. More roots in drought treatment were produced in the wetter soil horizons at 25-35 cm depth than at the surface. We conclude that fine and very fine root systems of Norway spruce have the capacity to re-allocate resources to roots at different depths in response to environmental signals, resulting in changes in necromass to biomass ratio.

Deviant disabilities: the exclusion of drug and alcohol addiction from the Equality Act 2010

The Equality Act 2010, in keeping with the Disability Discrimination Act 1995, excludes those identified as drug and alcohol ‘addicted’ from the scope of provisions prohibiting discrimination against disabled people. This article addresses the significance of, and justification for, this exclusion. It begins with a legislative background to the relevant limitation and subsequently examines its rationale according to prevailing legal, medical and sociological discourses. The article then considers the relevance of the discussion for disability rights. Although ‘addiction’, or the preferred term, ‘substance dependence’, is classified as a disability for international systems of disease classification, the relevance of substance dependence for discussion on disability rights, and of disability for discussion on substance dependence, has largely escaped critical comment.

Heterogeneous surface expression of EspA translocon filaments by Escherichia coli O157:H7 is controlled at the posttranscriptional level

Type III secretion systems of enteric bacteria enable translocation of effector proteins into host cells. Secreted proteins of verotoxigenic Escherichia coli O157 strains include components of a translocation apparatus, EspA, -B, and -D, as well as "effectors" such as the translocated intimin receptor (Tir) and the mitochondrion-associated protein (Map). This research has investigated the regulation of LEE4 translocon proteins, in particular EspA. EspA filaments could not be detected on the bacterial cell surface when E. coli O157:H7 was cultured in M9 minimal medium but were expressed from only a proportion of the bacterial population when cultured in minimal essential medium modified with 25 mM HEPES. The highest proportions of EspA-filamented bacteria were detected in late exponential phase, after which filaments were lost rapidly from the bacterial cell surface. Our previous research had shown that human and bovine E. coli O157:H7 strains exhibit marked differences in EspD secretion levels. Here it is demonstrated that the proportion of the bacterial population expressing EspA filaments was associated with the level of EspD secretion. The ability of individual bacteria to express EspA filaments was not controlled at the level of LEE1-4 operon transcription, as demonstrated by using both beta-galactosidase and green fluorescent protein (GFP) promoter fusions. All bacteria, whether expressing EspA filaments or not, showed equivalent levels of GFP expression when LEEI-4 translational fusions were used. Despite this, the LEE4-espADB mRNA was more abundant from populations with a high proportion of nonsecreting bacteria (low secretors) than from populations with a high proportion of secreting and therefore filamented bacteria (high secretors). This research demonstrates that while specific environmental conditions are required to induce LEEI-4 expression, a further checkpoint exists before EspA filaments are produced on the bacterial surface and secretion of effector proteins occurs. This checkpoint in E. coli O157:H7 translocon expression is controlled by a posttranscriptional mechanism acting on LEE4-espADB mRNA. The heterogeneity in EspA filamentation could arise from phase-variable expression of regulators that control this posttranscriptional mechanism.

One-port de-embedding technique for the quasi-optical characterization of integrated components

We describe a one-port de-embedding technique suitable for the quasi-optical characterization of terahertz integrated components at frequencies beyond the operational range of most vector network analyzers. This technique is also suitable when the manufacturing of precision terminations to sufficiently fine tolerances for the application of a TRL de-embedding technique is not possible. The technique is based on vector reflection measurements of a series of easily realizable test pieces. A theoretical analysis is presented for the precision of the technique when implemented using a quasi-optical null-balanced bridge reflectometer. The analysis takes into account quantization effects in the linear and angular encoders associated with the balancing procedure, as well as source power and detector noise equivalent power. The precision in measuring waveguide characteristic impedance and attenuation using this de-embedding technique is further analyzed after taking into account changes in the power coupled due to axial, rotational, and lateral alignment errors between the device under test and the instruments' test port. The analysis is based on the propagation of errors after assuming imperfect coupling of two fundamental Gaussian beams. The required precision in repositioning the samples at the instruments' test-port is discussed. Quasi-optical measurements using the de-embedding process for a WR-8 adjustable precision short at 125 GHz are presented. The de-embedding methodology may be extended to allow the determination of S-parameters of arbitrary two-port junctions. The measurement technique proposed should prove most useful above 325 GHz where there is a lack of measurement standards.

Satanic Mills? An illustration of Victorian external environmental accounting

The paper evaluates a Victorian environmental account of the pollution of the River Wandle. This account was produced during a period of social and environmental crisis, when there were no significant industrial environmental regulations. This problematising external environmental account provides valuable insights into the historical development of social and environmental accounting. Our analysis located this account within an institutional reform programme to create systems of governance to mitigate the damage arising from unfettered industrial growth. We argue that problematising external environmental accounting has a longer tradition than previously recognised in the literature and predates corporate social and environmental reporting.

Contributions of roots and rootstocks to sustainable, intensified crop production

Sustainable intensification is seen as the main route for meeting the world’s increasing demands for food and fibre. As demands mount for greater efficiency in the use of resources to achieve this goal, so the focus on roots and rootstocks and their role in acquiring water and nutrients, and overcoming pests and pathogens, is increasing. The purpose of this review is to explore some of the ways in which understanding root systems and their interactions with soils could contribute to the development of more sustainable systems of intensive production. Physical interactions with soil particles limit root growth if soils are dense, but root–soil contact is essential for optimal growth and uptake of water and nutrients. X-ray microtomography demonstrated that maize roots elongated more rapidly with increasing root–soil contact, as long as mechanical impedance was not limiting root elongation, while lupin was less sensitive to changes in root–soil contact. In addition to selecting for root architecture and rhizosphere properties, the growth of many plants in cultivated systems is profoundly affected by selection of an appropriate rootstock. Several mechanisms for scion control by rootstocks have been suggested, but the causal signals are still uncertain and may differ between crop species. Linkage map locations for quantitative trait loci for disease resistance and other traits of interest in rootstock breeding are becoming available. Designing root systems and rootstocks for specific environments is becoming a feasible target.

Scattering of electromagnetic waves by rough interfaces and inhomogeneous layers

We consider a two-dimensional problem of scattering of a time-harmonic electromagnetic plane wave by an infinite inhomogeneous conducting or dielectric layer at the interface between semi-infinite homogeneous dielectric half-spaces. The magnetic permeability is assumed to be a fixed positive constant. The material properties of the media are characterized completely by an index of refraction, which is a bounded measurable function in the layer and takes positive constant values above and below the layer, corresponding to the homogeneous dielectric media. In this paper, we examine only the transverse magnetic (TM) polarization case. A radiation condition appropriate for scattering by infinite rough surfaces is introduced, a generalization of the Rayleigh expansion condition for diffraction gratings. With the help of the radiation condition the problem is reformulated as an equivalent mixed system of boundary and domain integral equations, consisting of second-kind integral equations over the layer and interfaces within the layer. Assumptions on the variation of the index of refraction in the layer are then imposed which prove to be sufficient, together with the radiation condition, to prove uniqueness of solution and nonexistence of guided wave modes. Recent, general results on the solvability of systems of second kind integral equations on unbounded domains establish existence of solution and continuous dependence in a weighted norm of the solution on the given data. The results obtained apply to the case of scattering by a rough interface between two dielectric media and to many other practical configurations.