Special issue on advances in data mining and its applications

Data mining is the process to identify valid, implicit, previously unknown, potentially useful and understandable information from large databases. It is an important step in the process of knowledge discovery in databases, (Olaru & Wehenkel, 1999). In a data mining process, input data can be structured, seme-structured, or unstructured. Data can be in text, categorical or numerical values. One of the important characteristics of data mining is its ability to deal data with large volume, distributed, time variant, noisy, and high dimensionality. A large number of data mining algorithms have been developed for different applications. For example, association rules mining can be useful for market basket problems, clustering algorithms can be used to discover trends in unsupervised learning problems, classification algorithms can be applied in decision-making problems, and sequential and time series mining algorithms can be used in predicting events, fault detection, and other supervised learning problems (Vapnik, 1999). Classification is among the most important tasks in the data mining, particularly for data mining applications into engineering fields. Together with regression, classification is mainly for predictive modelling. So far, there have been a number of classification algorithms in practice. According to (Sebastiani, 2002), the main classification algorithms can be categorized as: decision tree and rule based approach such as C4.5 (Quinlan, 1996); probability methods such as Bayesian classifier (Lewis, 1998); on-line methods such as Winnow (Littlestone, 1988) and CVFDT (Hulten 2001), neural networks methods (Rumelhart, Hinton & Wiliams, 1986); example-based methods such as k-nearest neighbors (Duda & Hart, 1973), and SVM (Cortes & Vapnik, 1995). Other important techniques for classification tasks include Associative Classification (Liu et al, 1998) and Ensemble Classification (Tumer, 1996).

Quasistationarity of continuous-time Markov chains with positive drift

We shall study continuous-time Markov chains on the nonnegative integers which are both irreducible and transient, and which exhibit discernible stationarity before drift to infinity sets in. We will show how this 'quasi' stationary behaviour can be modelled using a limiting conditional distribution: specifically, the limiting state probabilities conditional on not having left 0 for the last time. By way of a dual chain, obtained by killing the original process on last exit from 0, we invoke the theory of quasistationarity for absorbing Markov chains. We prove that the conditioned state probabilities of the original chain are equal to the state probabilities of its dual conditioned on non-absorption, thus allowing us to establish the simultaneous existence and then equivalence, of their limiting conditional distributions. Although a limiting conditional distribution for the dual chain is always a quasistationary distribution in the usual sense, a similar statement is not possible for the original chain.

New methods for determining quasi-stationary distributions for Markov chains

We shall be concerned with the problem of determining quasi-stationary distributions for Markovian models directly from their transition rates Q. We shall present simple conditions for a mu-invariant measure m for Q to be mu-invariant for the transition function, so that if m is finite, it can be normalized to produce a quasi-stationary distribution. (C) 2000 Elsevier Science Ltd. All rights reserved.

The land assembly problem revisited

As in the standard land assembly problem, a developer wants to buy two adjacent blocks of land belonging to two different owners. The value of the two blocks of land to the developer is greater than the sum of the individual values of the blocks for each owner. Unlike the land assembly literature, however, our focus is on the incentive that each lot owner has to delay the start of negotiations, rather than on the public goods nature of the problem. An incentive for delay exists, for example, when owners perceive that being last to sell will allow them to capture a larger share of the joint surplus from the development. We show that competition at point of sale can cause equilibrium delay, and that cooperation at point of sale will eliminate delay. This suggests that strategic delay is another source for the inefficient allocation of land, in addition to the public-good type externality pointed out by Grossman and Hart [Bell Journal of Economics 11 (1980) 42] and O'Flaherty [Regional Science and Urban Economics 24 (1994) 287]. (C) 2004 Elsevier B.V. All rights reserved.

The Object Management Group Ontology Definition Metamodel

Report of a submission being made to a major international software engineering standards group, the Object Management Group which ties together OMG standards with World-Wide Web Consortium and International Standards Organization standards. Major industry bodies including IBM are collaborating, and the submission has the support of 24 companies. OMG, W3C and ISO standards strongly influence the industry, especially in combination. Colomb was a major contributor, responsible for 30% of the submission, and the primary author of the paper.

Variations in cone photoreceptor abundance and the visual ecology of birds

The relative abundance and topographical distribution of retinal cone photoreceptors was measured in 19 bird species to identify possible correlations between photoreceptor complement and visual ecology. In contrast to previous studies, all five types of cone photoreceptor were distinguished, using bright field and epifluorescent light microscopy, in four retinal quadrants. Land birds tended to show either posterior dorsal to anterior ventral or anterior dorsal to posterior ventral gradients in cone photoreceptor distribution, fundus coloration and oil droplet pigmentation across the retina. Marine birds tended to show dorsal to ventral gradients instead. Statistical analyses showed that the proportions of the different cone types varied significantly across the retinae of all species investigated. Cluster analysis was performed on the data to identify groups or clusters of species on the basis of their oil droplet complement. Using the absolute percentages of each oil droplet type in each quadrant for the analysis produced clusters that tended to reflect phylogenetic relatedness between species rather than similarities in their visual ecology. Repeating the analysis after subtracting the mean percentage of a given oil droplet type across the whole retina (the 'eye mean') from the percentage of that oil droplet type in each quadrant, i.e. to give a measure of the variation about the mean, resulted in clusters that reflected diet, feeding behaviour and habitat to a greater extent than phylogeny.

The visual ecology of avian photoreceptors

The spectral sensitivities of avian retinal photoreceptors are examined with respect to microspectrophotometric measurements of single cells, spectrophotometric measurements of extracted or in vitro regenerated visual pigments, and molecular genetic analyses of visual pigment opsin protein sequences. Bird species from diverse orders are compared in relation to their evolution, their habitats and the multiplicity of visual tasks they must perform. Birds have five different types of visual pigment and seven different types of photo receptor-rods, double (uneven twin) cones and four types of single cone. The spectral locations of the wavelengths of maximum absorbance (lambda (max)) of the different visual pigments, and the spectral transmittance characteristics of the intraocular spectral filters (cone oil droplets) that also determine photoreceptor spectral sensitivity, vary according to both habitat and phylogenetic relatedness. The primary influence on avian retinal design appears to be the range of wavelengths available for vision, regardless of whether that range is determined by the spectral distribution of the natural illumination or the spectral transmittance of the ocular media (cornea, aqueous humour, lens, vitreous humour). Nevertheless, other variations in spectral sensitivity exist that reflect the variability and complexity of avian visual ecology. (C) 2001 Elsevier Science Ltd. All rights reserved.

CD40 and CD86 upregulation with divergent CMRF44 expression on blood dendritic cells in inflammatory bowel diseases

OBJECTIVE: Dendritic cells (DC) are the only antigen-presenting cells that can activate naive T lymphocytes and initiate a primary immune response. They are also thought to have a role in immune tolerance. DC traffic from the blood to peripheral tissue where they become activated. They then present antigen and the costimulating signals necessary to initiate an immune response. In this study, we investigated the number, subsets, and activation pattern of circulating and intestinal DC from patients with clinically mild ulcerative colitis (UC) or Crohn's disease. METHODS: Patients were recruited, if they were not taking immunosuppressive therapy, and were assessed for clinical severity of their disease using for UC, the Clinical Activity Index, and for Crohn's disease, the Crohn's Disease Activity Index. Blood CD11c(+) and CD11c(-) DC subsets, expression of costimulatory antigens, CD86 and CD40, and the early differentiation/activation antigen, CMRF44, were enumerated by multicolor flow cytometry of lineage negative (lin(-) = CD3(-), CD19(-), CD14(-), CD16(-)) HLA-DR+ DC. These data were compared with age-matched healthy and the disease control groups of chronic noninflammatory GI diseases (cGI), acute noninflammatory GI diseases (aGI), and chronic non-GI inflammation (non-GI). In addition, cryostat sections of colonoscopic biopsies from healthy control patients and inflamed versus noninflamed gut mucosa of inflammatory bowel disease (IBD) patients were examined for CD86(+) and CD40(+)lin(-) cells. RESULTS: Twenty-one Crohn's disease and 25 UC patients, with mean Crohn's Disease Activity Index of 98 and Clinical Activity Index of 3.1, and 56 healthy controls, five cGI, five aGI, and six non-GI were studied. CD11c(+) and CD11c(-) DC subsets did not differ significantly between Crohn's, UC, and healthy control groups. Expression of CD86 and CD40 on freshly isolated blood DC from Crohn's patients appeared higher (16.6%, 31%) and was significantly higher in UC (26.6%, 46.3%) versus healthy controls (5.5%, 25%) (p = 0.004, p = 0.012) and non-GI controls (10.2%, 22.8%) (p = 0.012, p = 0.008), but not versus cGI or aGI controls. CD86(+) and CD40(+) DC were also present in inflamed colonic and ileal mucosa from UC and Crohn's patients but not in noninflamed IBD mucosa or normal mucosa. Expression of the CMRF44 antigen was low on freshly isolated DC, but it was upregulated after 24-h culture on DC from all groups, although significantly less so on DC from UC versus Crohn's or healthy controls (p = 0.024). The CMRF44(+) antigen was mainly associated with CD11c(+) DC, and in UC was inversely related to the Clinical Activity Index (r = -0.69, p = 0.0002). CONCLUSIONS: There is upregulation of costimulatory molecules on blood DC even in very mild IBD but surprisingly, there is divergent expression of the differentiation/activation CMRF44 antigen. Upregulation of costimulatory molecules and divergent expression of CMRF44 in blood DC was also apparent in cGI and aGI but not in non-GI or healthy controls, whereas intestinal CD86(+) and CD40(+) DC were found only in inflamed mucosa from IBD patients. Persistent or distorted activation of blood DC or divergent regulation of costimulatory and activation antigens may have important implications for gut mucosal immunity and inflammation. (Am J Gastroenterol 2001;96:2946-2956. (C) 2001 by Am. Coll. of Gastroenterology).

Slagging propensities of blended coals

This paper presents the experimental results on the slagging propensity of three pairs of blended coals tested in the Australian Coal Industry Research Laboratory furnace. The results showed that none of the coals or blends produced strongly molten deposit. At worst, the deposits contained some moderately sintered material. Most of the blends have (slightly) worse slagging behavior than the component coals. In order to rank the slagging propensity numerically, we defined the minimum heat flux ratio and total heat flux ratio based on the heat flux profiles. They are better in ranking the slagging propensities than other measures such as the build-up rate and visual physical characteristics. The Fe2O3/CaO molar ratio correlates the slagging propensity for five coals and six blends of this study. The worst slagging occurs when the ratio approaches 1.0. This ratio provides explanation of why the blends had worse slagging than the component coals for the pairs of blends: A-B and C-D. However, we note that there are causes of slagging other than the Fe2O3/CaO molar ratio. (C) 2001 Elsevier Science Ltd. All rights reserved.