Machine learning algorithms for spatial data. Case studies: Environmental pollution, natural hazards, renewable resources

This paper presents general problems and approaches for the spatial data analysis using machine learning algorithms. Machine learning is a very powerful approach to adaptive data analysis, modelling and visualisation. The key feature of the machine learning algorithms is that they learn from empirical data and can be used in cases when the modelled environmental phenomena are hidden, nonlinear, noisy and highly variable in space and in time. Most of the machines learning algorithms are universal and adaptive modelling tools developed to solve basic problems of learning from data: classification/pattern recognition, regression/mapping and probability density modelling. In the present report some of the widely used machine learning algorithms, namely artificial neural networks (ANN) of different architectures and Support Vector Machines (SVM), are adapted to the problems of the analysis and modelling of geo-spatial data. Machine learning algorithms have an important advantage over traditional models of spatial statistics when problems are considered in a high dimensional geo-feature spaces, when the dimension of space exceeds 5. Such features are usually generated, for example, from digital elevation models, remote sensing images, etc. An important extension of models concerns considering of real space constrains like geomorphology, networks, and other natural structures. Recent developments in semi-supervised learning can improve modelling of environmental phenomena taking into account on geo-manifolds. An important part of the study deals with the analysis of relevant variables and models' inputs. This problem is approached by using different feature selection/feature extraction nonlinear tools. To demonstrate the application of machine learning algorithms several interesting case studies are considered: digital soil mapping using SVM, automatic mapping of soil and water system pollution using ANN; natural hazards risk analysis (avalanches, landslides), assessments of renewable resources (wind fields) with SVM and ANN models, etc. The dimensionality of spaces considered varies from 2 to more than 30. Figures 1, 2, 3 demonstrate some results of the studies and their outputs. Finally, the results of environmental mapping are discussed and compared with traditional models of geostatistics.

Differentiation of blue ballpoint pen by positive and negative mode LDI-MS

Usually, the differentiation of inks on questioned documents is carried out by optical methods and thin layer chromatography (TLC). Therefore, spectrometric methods were also proposed in forensic literature for the analysis of dyes. Between these techniques, laser desorption/ionization mass spectrometry (LDI-MS) has demonstrated a great versatility thanks to its sensitivity to blue ballpoint ink dyes and minimal sample destruction. Previous researches concentrated mostly on the LDI-MS positive mode and have shown that this analytical tool offers higher discrimination power than high performance TLC (HPTLC) for the differentiation of blue ballpoint inks. Although LDI-MS negative mode has already been applied in numerous forensic domains like the studies of works of art, automotive paints or rollerball pens, its potential for the discrimination of ballpoint pens was never studied before. The aim of the present paper is therefore to evaluate its potential for the discrimination of blue ballpoint inks. After optimization of the method, ink entries from 33 blue ballpoint pens were analyzed directly on paper in both positive and negative modes by LDI-MS. Several cationic and anionic ink components were identified in inks; therefore, pens were classified and compared according to their formulations. Results show that additional information provided by anionic dyes and pigments significantly increases the discrimination power of positive mode. In fact, it was demonstrated that classifications obtained by the two modes were, to some extent, complementary (i.e., inks with specific cationic dyes not necessarily contained the same anionic components).

Molecular differentiation of Angiostrongylus costaricensis, A. cantonensis, and A. vasorum by polymerase chain reaction- restriction fragment length polymorphism

Angiostrongylus cantonensis, A. costaricensis, and A. vasorum are etiologic agents of human parasitic diseases. Their identification, at present, is only possible by examining the adult worm after a 40-day period following infection of vertebrate hosts with the third-stage larvae. In order to obtain a diagnostic tool to differentiate larvae and adult worm from the three referred species, polymerase chain reaction-restriction fragment length polymorphism was carried out. The rDNA second internal transcribed spacer (ITS2) and mtDNA cytochrome oxidase I regions were amplified, followed by digestion of fragments with the restriction enzymes RsaI, HapII, AluI, HaeIII, DdeI and ClaI. The enzymes RsaI and ClaI exhibited the most discriminating profiles for the differentiation of the regions COI of mtDNA and ITS2 of rDNA respectively. The methodology using such regions proved to be efficient for the specific differentiation of the three species of Angiostrongylus under study.

Trace element differentiation in ferruginous accumulation soil patterns under tropical rainforest of southern Cameroon, the role of climatic change

Regions under tropical rainforest cover, such as central Africa and Brazil are characterised by degradation and dismantling of old ferricrete structures. In southern Cameroon, these processes are relayed by present-day ferruginous accumulation soil facies, situated on the middle and the lower part of hill slopes. These facies become progressively harder towards the surface, containing from bottom to top, mainly kaolinite, kaolinite-goethite and Al-rich goethite-hematite, and are discontinuous to the relictic hematite-dominated ferricrete that exist in the upper part of the hill slope. These features were investigated in terms of geochemical differentiation of trace elements. It appears that, in contrast to the old ferricrete facies, the current ferruginous accumulations are enriched in transitional trace elements (V, Cr, Co, Y, Sc) and Ph, while alkali-earth elements are less differentiated. This recent chemical accumulation is controlled both by intense weathering of the granodiorite bedrock and by mobilisation of elements previously accumulated in the old ferricrete. The observed processes are clearly linked to the present-day humid climate with rising groundwater tables. They slowly replace the old ferricretes formed during Cretaceous time under more seasonal climatic conditions, representing an instructive case of continuos global change. (C) 2002 Elsevier Science B.V. All rights reserved.

Suppression of Arabidopsis protophloem differentiation and root meristem growth by CLE45 requires the receptor-like kinase BAM3.

Peptide signaling presumably occupies a central role in plant development, yet only few concrete examples of receptor-ligand pairs that act in the context of specific differentiation processes have been described. Here we report that second-site null mutations in the Arabidopsis leucine-rich repeat receptor-like kinase gene barely any meristem 3 (BAM3) perfectly suppress the postembryonic root meristem growth defect and the associated perturbed protophloem development of the brevis radix (brx) mutant. The roots of bam3 mutants specifically resist growth inhibition by the CLAVATA3/ENDOSPERM SURROUNDING REGION 45 (CLE45) peptide ligand. WT plants transformed with a construct for ectopic overexpression of CLE45 could not be recovered, with the exception of a single severely dwarfed and sterile plant that eventually died. By contrast, we obtained numerous transgenic bam3 mutants transformed with the same construct. These transgenic plants displayed a WT phenotype, however, supporting the notion that CLE45 is the likely BAM3 ligand. The results correlate with the observation that external CLE45 application represses protophloem differentiation in WT, but not in bam3 mutants. BAM3, BRX, and CLE45 are expressed in a similar spatiotemporal trend along the developing protophloem, up to the end of the transition zone. Induction of BAM3 expression upon CLE45 application, ectopic overexpression of BAM3 in brx root meristems, and laser ablation experiments suggest that intertwined regulatory activity of BRX, BAM3, and CLE45 could be involved in the proper transition of protophloem cells from proliferation to differentiation, thereby impinging on postembryonic growth capacity of the root meristem.

Spatial analysis of the distribution of leprosy in the State of Ceará, Northeast Brazil

The aim of this study was to describe spatial patterns of the distribution of leprosy and to investigate spatial clustering of incidence rates in the state of Ceará, Northeast Brazil. The average incidence rate of leprosy for the period of 1991 to 1999 was calculated for each municipality of Ceará. Maps were used to describe the spatial distribution of the disease, and spatial statistics were applied to explore large- and small-scale variations of incidence rates. Three regions were identified in which the incidence of leprosy was particularly high. A spatial gradient in the incidence rates was identified, with a tendency of high rates to be concentrated on the North-South axis in the middle region of the state. Moran's I statistic indicated that a significant spatial autocorrelation also existed. The spatial distribution of leprosy in Ceará is heterogeneous. The reasons for spatial clustering of disease rates are not known, but might be related to an heterogeneous distribution of other factors such as crowding, social inequality, and environmental characteristics which by themselves determine the transmission of Mycobacterium leprae.

Mycobacterium tuberculosis complex differentiation using gyrB-restriction fragment length polymorphism analysis

Mycobacterium tuberculosis complex (MTBC) members are causative agents of human and animal tuberculosis. Differentiation of MTBC members is required for appropriate treatment of individual patients and for epidemiological purposes. Strains from six MTBC species - M. tuberculosis, M. bovis subsp. bovis, M. bovis BCG, M. africanum, M. pinnipedii, and "M. canetti" - were studied using gyrB-restriction fragment length polymorphism (gyrB-RFLP) analysis. A table was elaborated, based on observed restriction patterns and published gyrB sequences. To evaluate applicability of gyrB-RFLP at Instituto Adolfo Lutz, São Paulo, Mycobacterial Reference Laboratory, 311 MTBC clinical isolates, previously identified using traditional methods as M. tuberculosis (306), M. bovis (3), and M. bovis BCG (2), were analyzed by gyrB-RFLP. All isolates were correctly identified by the molecular method, but no distinction between M. bovis and M. bovis BCG was obtained. Differentiation of M. tuberculosis and M. bovis is of utmost importance, because they require different treatment schedules. In conclusion, gyrB-RFLP is accurate and easy-to-perform, with potential to reduce time needed for conventional differentiation methods. However, application for epidemiological studies remains limited, because it cannot differentiate M. tuberculosis from M. africanum subtype II, and "M. canetti", M. africanum subtype I from M. pinnipedii, and. M. bovis from M. bovis BCG.

Spatial distribution of schistosomiasis foci on Itamaracá Island, Pernambuco, Brazil

Acute cases of schistosomiasis have been found on the coastal area of Pernambuco, Brazil, due to environmental disturbances and disorderly occupation of the urban areas. This study identifies and spatially marks the main foci of the snail host species, Biomphalaria glabrata on Itamaracá Island. The chaotic occupation of the beach resorts has favoured the emergence of transmission foci, thus exposing residents and tourists to the risk of infection. A database covering five years of epidemiological investigation on snails infected by Schistosoma mansoni in the island was produced with information from the geographic positioning of the foci, number of snails collected, number of snails tested positive, and their infection rate. The spatial position of the foci were recorded through the Global Positioning System (GPS), and the geographical coordinates were imported by AutoCad. The software packages ArcView and Spring were used for data processing and spatial analysis. AutoCad 2000 was used to plot the pairs of coordinates obtained from GPS. Between 1998 and 2002 5009 snails, of which 12.2% were positive for S. mansoni, were collected in Forte Beach. A total of 27 foci and areas of environmental risk were identified and spatially analyzed allowing the identification of the areas exposed to varying degrees of risk.

Protein kinase C-activating tumor promoters enhance the differentiation of astrocytes in aggregating fetal brain cell cultures.

Serum-free aggregating cell cultures of fetal rat telencephalon treated with the potent tumor promoter phorbol 12-myristate 13-acetate (PMA) showed a marked, rapid, and sustained increase in the activity of the astrocyte-specific enzyme glutamine synthetase (GS). This effect was accompanied by a small increase in RNA synthesis and a progressive reduction in DNA synthesis. Only mitotically active cultures were responsive to PMA treatments. Since in aggregate cultures astrocytes are the preponderant cell type, both in number and mitotic activity, it can be concluded that PMA induces and/or enhances the terminal differentiation of astrocytes. The developmental expression of GS was also greatly stimulated by mezerein, a potent nonphorbol tumor promoter, but not by 4 alpha-phorbol 12,13-didecanoate, a nonpromoting phorbol ester. Since both tumor promoters, PMA and mezerein, are potent and specific activators of C-kinase, it is suggested that C-kinase plays a regulatory role in the growth and differentiation of normal astrocytes.

Molecular differentiation of Anopheles (Nyssorhynchus) benarrochi and An. (N.) oswaldoi from Southern Colombia

Anopheles (Nyssorhynchus) benarrochi, An. (N.) oswaldoi, and An. (N.) rangeli are the most common anthropophilic mosquitoes in the southern Colombian state of Putumayo. Adult females are most commonly collected in epidemiological studies, and this stage poses significant problems for correct identification, due to overlapping inter-specific morphological characters. Although An. rangeli is easy to identify, the morphological variant of An. benarrochi found in the region and An. oswaldoi are not always easy to separate. Herein we provide a rapid molecular method to distinguish these two species in Southern Colombia. Sequence data for the second internal transcribed spacer (ITS2) region of rDNA was generated for link-reared progeny of An. benarrochi and An. oswaldoi, that had been identified using all life stages. ITS2 sequences were 540 bp in length in An. benarrochi (n = 9) and 531 bp in An. oswaldoi (n = 7). Sequences showed no intra-specific variation and ungapped inter-specific sequence divergence was 6.4%. Species diagnostic banding patterns were recovered following digestion of the ITS2 amplicons with the enzyme Hae III as follows: An. benarrochi (365, 137, and 38 bp) and An. oswaldoi (493 and 38 bp). This polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) assay provides rapid, accurate, and inexpensive species diagnosis of adult females. This will benefit future epidemiological studies and, as PCR amplification can be achieved using a single mosquito leg, the remaining specimen can be either retained as a morphological voucher or further used in vector incrimination studies. That An. benarrochi comprises a complex of at least two species across Latin America is discussed.

Hiding the evidence: The State and spatial inequalities in health in Ireland

The Irish State has consistently reduced its production of publicly accessible disease / mortality maps over the last fifty years. State health statistics, and the small number of disease / mortality maps that have been produced in official publications, show a declining level of detail and are routinely out of date. Following a review of the production of disease / mortality maps in Ireland by the State and allied health agencies, two reasons are suggested for this decline. The first explanation relates to spatial inequalities in healthcare provision and to the absence of a health funding formulae in Ireland. The second explanation focuses on the potential politicisation of spatial inequalities in health status. Researchers in these fields are urged to disseminate information widely on spatial inequalities in healthcare provision, healthcare access and health status, both within and outside of the academic literature. Researchers in these fields are also urged to adopt an advocacy role on these issues, or to develop strategic alliances with such advocates.����