Estimating the number of clusters in multivariate data by self-organizing maps

Determining the structure of data without prior knowledge of the number of clusters or any information about their composition is a problem of interest in many fields, such as image analysis, astrophysics, biology, etc. Partitioning a set of n patterns in a p-dimensional feature space must be done such that those in a given cluster are more similar to each other than the rest. As there are approximately Kn/K! possible ways of partitioning the patterns among K clusters, finding the best solution is very hard when n is large. The search space is increased when we have no a priori number of partitions. Although the self-organizing feature map (SOM) can be used to visualize clusters, the automation of knowledge discovery by SOM is a difficult task. This paper proposes region-based image processing methods to post-processing the U-matrix obtained after the unsupervised learning performed by SOM. Mathematical morphology is applied to identify regions of neurons that are similar. The number of regions and their labels are automatically found and they are related to the number of clusters in a multivariate data set. New data can be classified by labeling it according to the best match neuron. Simulations using data sets drawn from finite mixtures of p-variate normal densities are presented as well as related advantages and drawbacks of the method.     

SOMCD: method for evaluating protein secondary structure from UV circular dichroism spectra

This article presents SOMCD, an improved method for the evaluation of protein secondary structure from circular dichroism spectra, based on Kohonen's self-organizing maps (SOM). Protein circular dichroism (CD) spectra are used to train a SOM, which arranges the spectra on a two-dimensional map. Location in the map reflects the secondary structure composition of a protein. With SOMCD, the prediction of beta-turn has been included. The number of spectra in the training set has been increased, and it now includes 39 protein spectra and 6 reference spectra. Finally, SOM parameters have been chosen to minimize distortion and make the network produce clusters with known properties. Estimation results show improvements compared with the previous version, K2D, which, in addition, estimated only three secondary structure components; the accuracy of the method is more uniform over the different secondary structures.     

Probing the Topological Properties of Complex Networks Modeling Short Written Texts

In recent years, graph theory has been widely employed to probe several language properties. More specifically, the so-called word adjacency model has been proven useful for tackling several practical problems, especially those relying on textual stylistic analysis. The most common approach to treat texts as networks has simply considered either large pieces of texts or entire books. This approach has certainly worked well—many informative discoveries have been made this way—but it raises an uncomfortable question: could there be important topological patterns in small pieces of texts? To address this problem, the topological properties of subtexts sampled from entire books was probed. Statistical analyses performed on a dataset comprising 50 novels revealed that most of the traditional topological measurements are stable for short subtexts. When the performance of the authorship recognition task was analyzed, it was found that a proper sampling yields a discriminability similar to the one found with full texts. Surprisingly, the support vector machine classification based on the characterization of short texts outperformed the one performed with entire books. These findings suggest that a local topological analysis of large documents might improve its global characterization. Most importantly, it was verified, as a proof of principle, that short texts can be analyzed with the methods and concepts of complex networks. As a consequence, the techniques described here can be extended in a straightforward fashion to analyze texts as time-varying complex networks.     

Plants mediate soil organic matter decomposition in response to sea level rise

Tidal marshes have a large capacity for producing and storing organic matter, making their role in the global carbon budget disproportionate to land area. Most of the organic matter stored in these systems is in soils where it contributes 2–5 times more to surface accretion than an equal mass of minerals. Soil organic matter (SOM) sequestration is the primary process by which tidal marshes become perched high in the tidal frame, decreasing their vulnerability to accelerated relative sea level rise (RSLR). Plant growth responses to RSLR are well understood and represented in century‐scale forecast models of soil surface elevation change. We understand far less about the response of SOM decomposition to accelerated RSLR. Here we quantified the effects of flooding depth and duration on SOM decomposition by exposing planted and unplanted field‐based mesocosms to experimentally manipulated relative sea level over two consecutive growing seasons. SOM decomposition was quantified as CO₂ efflux, with plant‐ and SOM‐derived CO₂ separated via δ¹³CO₂. Despite the dominant paradigm that decomposition rates are inversely related to flooding, SOM decomposition in the absence of plants was not sensitive to flooding depth and duration. The presence of plants had a dramatic effect on SOM decomposition, increasing SOM‐derived CO₂ flux by up to 267% and 125% (in 2012 and 2013, respectively) compared to unplanted controls in the two growing seasons. Furthermore, plant stimulation of SOM decomposition was strongly and positively related to plant biomass and in particular aboveground biomass. We conclude that SOM decomposition rates are not directly driven by relative sea level and its effect on oxygen diffusion through soil, but indirectly by plant responses to relative sea level. If this result applies more generally to tidal wetlands, it has important implications for models of SOM accumulation and surface elevation change in response to accelerated RSLR.     

Sensitivity of mangrove soil organic matter decay to warming and sea level change

Mangroves are among the world's most carbon‐dense ecosystems, but they are threatened by rapid climate change and rising sea levels. The accumulation and decomposition of soil organic matter (SOM) are closely tied to mangroves' carbon sink functions and resistance to rising sea levels. However, few studies have investigated the response of mangrove SOM dynamics to likely future environmental conditions. We quantified how mangrove SOM decay is affected by predicted global warming (+4°C), sea level changes (simulated by altering of the inundation duration to 0, 2, and 6 hr/day), and their interaction. Whilst changes in inundation duration between 2 and 6 hr/day did not affect SOM decay, the treatment without inundation led to a 60% increase. A warming of 4°C caused SOM decay to increase by 21%, but longer inundation moderated this temperature‐driven increase. Our results indicate that (a) sea level rise is unlikely to decrease the SOM decay rate, suggesting that previous mangrove elevation gain, which has allowed mangroves to persist in areas of sea level rise, might result from changes in root production and/or mineral sedimentation; (b) sea level fall events, predicted to double in frequency and area, will cause periods of intensified SOM decay; (c) changing tidal regimes in mangroves due to sea level rise might attenuate increases in SOM decay caused by global warming. Our results have important implications for forecasting mangrove carbon dynamics and the persistence of mangroves and other coastal wetlands under future scenarios of climate change.     

Rotation-Invariant Features for Multi-Oriented Text Detection in Natural Images

Texts in natural scenes carry rich semantic information, which can be used to assist a wide range of applications, such as object recognition, image/video retrieval, mapping/navigation, and human computer interaction. However, most existing systems are designed to detect and recognize horizontal (or near-horizontal) texts. Due to the increasing popularity of mobile-computing devices and applications, detecting texts of varying orientations from natural images under less controlled conditions has become an important but challenging task. In this paper, we propose a new algorithm to detect texts of varying orientations. Our algorithm is based on a two-level classification scheme and two sets of features specially designed for capturing the intrinsic characteristics of texts. To better evaluate the proposed method and compare it with the competing algorithms, we generate a comprehensive dataset with various types of texts in diverse real-world scenes. We also propose a new evaluation protocol, which is more suitable for benchmarking algorithms for detecting texts in varying orientations. Experiments on benchmark datasets demonstrate that our system compares favorably with the state-of-the-art algorithms when handling horizontal texts and achieves significantly enhanced performance on variant texts in complex natural scenes.     

土壤有机质概念和分组技术研究进展

土壤有机质一直是土壤学研究领域的重点,在过去的50年里,对土壤质量可持续性观念的增强和寻找快速判断人为因素对土壤质量影响方向指标的强烈愿望导致了土壤有机质的研究重点发生了急剧变化：对农业措施反映慢的土壤腐殖质类物质的研究正在退出土壤有机质研究领域,而侧重点逐渐转向了土壤中未受微生物作用或正在受微生物降解的有机残体;也出现了新的土壤有机质研究概念和对应测试手段：土壤有机质的比重分组、与有机质结合的土壤颗粒大小分组、土壤团聚体中的POM和iPOM以及土壤水溶性有机质和微生物体C等概念和测试手段被相继提了出来,土壤有机质的研究重点正在从土壤微生物的作用产物(腐殖质)向土壤微生物作用前的、具有部分生物活性的有机质(轻组有机质、砂粒组和粗粉砂粒组中的有机质、POM和iPOM)和完全具有生物活性的有机质(微生物体C和水溶性有机质)转移,这一过程与土壤有机质概念的拓展密不可分。     

Conceptualizing soil organic matter into particulate and mineral‐associated forms to address global change in the 21st century

Managing soil organic matter (SOM) stocks to address global change challenges requires well‐substantiated knowledge of SOM behavior that can be clearly communicated between scientists, management practitioners, and policy makers. However, SOM is incredibly complex and requires separation into multiple components with contrasting behavior in order to study and predict its dynamics. Numerous diverse SOM separation schemes are currently used, making cross‐study comparisons difficult and hindering broad‐scale generalizations. Here, we recommend separating SOM into particulate (POM) and mineral‐associated (MAOM) forms, two SOM components that are fundamentally different in terms of their formation, persistence, and functioning. We provide evidence of their highly contrasting physical and chemical properties, mean residence times in soil, and responses to land use change, plant litter inputs, warming, CO₂ enrichment, and N fertilization. Conceptualizing SOM into POM versus MAOM is a feasible, well‐supported, and useful framework that will allow scientists to move beyond studies of bulk SOM, but also use a consistent separation scheme across studies. Ultimately, we propose the POM versus MAOM framework as the best way forward to understand and predict broad‐scale SOM dynamics in the context of global change challenges and provide necessary recommendations to managers and policy makers.     

Exact solutions to the continuous-quality equation for soil organic matter turnover

All living systems depend on transformations of elements between different states. In particular, the transformation of dead organic matter in the soil (SOM) by decomposers (microbes) releases elements incorporated in SOM and makes the elements available anew to plants. A major problem in analysing and describing this process is that SOM, as the result of the decomposer activity, is a mixture of a very large number of molecules with widely differing chemical and physical properties. The continuous-quality equation (CQE) is a general equation describing this complexity by assigning a continuous-quality variable to each carbon atom in SOM. The use of CQE has been impeded by its complicated mathematics. Here, we show by deriving exact solutions that, at least for some specific cases, there exist solutions to CQE. These exact solutions show that previous approximations have overestimated the rate by which litter decomposes and as a consequence underestimated steady state SOM amounts. The exact and approximate solutions also differ with respect to the parameter space in which they yield finite steady-state SOM amounts. The latter point is important because temperature is one of the parameters and climatic change may move the solution from a region of the parameter space with infinite steady-state SOM to a region of finite steady-state SOM, with potentially large changes in soil carbon stores. We also show that the solution satisfies the Chapman-Kolmogorov theorem. The importance of this is that it provides efficient algorithms for numerical solutions.     

Pyrolysis-molecular beam mass spectrometry to characterize soil organic matter composition in chemically isolated fractions from differing land uses

Today’s questions concerning the role of soil organic matter (SOM) in soil fertility, ecosystem functioning and global change can only be addressed through knowledge of the controls on SOM stabilization and their interactions. Pyrolysis molecular beam mass spectrometry (py-MBMS) provides a powerful and rapid means of assessing the biochemical composition of SOM. However, characterization of SOM composition alone is insufficient to predict its dynamic behavior. Chemical fractionation is frequently used to isolate more homogeneous SOM components, but the composition of fractions is frequently unknown. We characterized biochemical SOM composition in two previously studied soils from the USA, under contrasting land uses: cultivated agriculture and native vegetation. Bulk soils, as well as chemically isolated SOM fractions (humic acid, humin and non-acid hydrolysable), were analyzed using py-MBMS. Principal components analysis (PCA) showed distinct differences in the SOM composition of isolated fractions. Py-MBMS spectra and PCA loadings were dominated by low molecular weight fragments associated with peptides and other N-containing compounds. The py-MBMS spectra were similar for native whole-soil samples under different vegetation, while cultivation increased heterogeneity. An approach based on previously published data on marker signals also suggests the importance of peptides in distinguishing samples. While the approach described here represents significant progress in the characterization of changing SOM composition, a truly quantitative analysis will only be achieved using multiple internal standards and by correcting for inorganic interference during py-MBMS analysis. Overall, we have provided proof of principle that py-MBMS can be a powerful tool to understand the controls on SOM dynamics, and further method development is underway.     

Self-Organising Maps and Correlation Analysis as a Tool to Explore Patterns in Excitation-Emission Matrix Data Sets and to Discriminate Dissolved Organic Matter Fluorescence Components

Dissolved organic matter (DOM) is a complex mixture of organic compounds, ubiquitous in marine and freshwater systems. Fluorescence spectroscopy, by means of Excitation-Emission Matrices (EEM), has become an indispensable tool to study DOM sources, transport and fate in aquatic ecosystems. However the statistical treatment of large and heterogeneous EEM data sets still represents an important challenge for biogeochemists. Recently, Self-Organising Maps (SOM) has been proposed as a tool to explore patterns in large EEM data sets. SOM is a pattern recognition method which clusterizes and reduces the dimensionality of input EEMs without relying on any assumption about the data structure. In this paper, we show how SOM, coupled with a correlation analysis of the component planes, can be used both to explore patterns among samples, as well as to identify individual fluorescence components. We analysed a large and heterogeneous EEM data set, including samples from a river catchment collected under a range of hydrological conditions, along a 60-km downstream gradient, and under the influence of different degrees of anthropogenic impact. According to our results, chemical industry effluents appeared to have unique and distinctive spectral characteristics. On the other hand, river samples collected under flash flood conditions showed homogeneous EEM shapes. The correlation analysis of the component planes suggested the presence of four fluorescence components, consistent with DOM components previously described in the literature. A remarkable strength of this methodology was that outlier samples appeared naturally integrated in the analysis. We conclude that SOM coupled with a correlation analysis procedure is a promising tool for studying large and heterogeneous EEM data sets.     

Investigating heterogeneous protein annotations toward cross-corpora utilization

Background  

The number of corpora, collections of structured texts, has been increasing, as a result of the growing interest in the application of natural language processing methods to biological texts. Many named entity recognition (NER) systems have been developed based on these corpora. However, in the biomedical community, there is yet no general consensus regarding named entity annotation; thus, the resources are largely incompatible, and it is difficult to compare the performance of systems developed on resources that were divergently annotated. On the other hand, from a practical application perspective, it is desirable to utilize as many existing annotated resources as possible, because annotation is costly. Thus, it becomes a task of interest to integrate the heterogeneous annotations in these resources.     

Assessment of self-organizing maps to analyze sole-carbon source utilization profiles

The use of community-level physiological profiles obtained with Biolog microplates is widely employed to consider the functional diversity of bacterial communities. Biolog produces a great amount of data which analysis has been the subject of many studies. In most cases, after some transformations, these data were investigated with classical multivariate analyses. Here we provided an alternative to this method, that is the use of an artificial intelligence technique, the Self-Organizing Maps (SOM, unsupervised neural network). We used data from a microcosm study of algae-associated bacterial communities placed in various nutritive conditions. Analyses were carried out on the net absorbances at two incubation times for each substrates and on the chemical guild categorization of the total bacterial activity. Compared to Principal Components Analysis and cluster analysis, SOM appeared as a valuable tool for community classification, and to establish clear relationships between clusters of bacterial communities and sole-carbon sources utilization. Specifically, SOM offered a clear bidimensional projection of a relatively large volume of data and were easier to interpret than plots commonly obtained with multivariate analyses. They would be recommended to pattern the temporal evolution of communities' functional diversity.     

An efficient approach to the travelling salesman problem using self-organizing maps

This paper presents an approach to the well-known Travelling Salesman Problem (TSP) using Self-Organizing Maps (SOM). The SOM algorithm has interesting topological information about its neurons configuration on cartesian space, which can be used to solve optimization problems. Aspects of initialization, parameters adaptation, and complexity analysis of the proposed SOM based algorithm are discussed. The results show an average deviation of 3.7% from the optimal tour length for a set of 12 TSP instances.     

Soil organic matter turnover is governed by accessibility not recalcitrance

Mechanisms to mitigate global climate change by sequestering carbon (C) in different ‘sinks' have been proposed as at least temporary measures. Of the major global C pools, terrestrial ecosystems hold the potential to capture and store substantially increased volumes of C in soil organic matter (SOM) through changes in management that are also of benefit to the multitude of ecosystem services that soils provide. This potential can only be realized by determining the amount of SOM stored in soils now, with subsequent quantification of how this is affected by management strategies intended to increase SOM concentrations, and used in soil C models for the prediction of the roles of soils in future climate change. An apparently obvious method to increase C stocks in soils is to augment the soil C pools with the longest mean residence times (MRT). Computer simulation models of soil C dynamics, e.g. RothC and Century, partition these refractory constituents into slow and passive pools with MRTs of centuries to millennia. This partitioning is assumed to reflect: (i) the average biomolecular properties of SOM in the pools with reference to their source in plant litter, (ii) the accessibility of the SOM to decomposer organisms or catalytic enzymes, or (iii) constraints imposed on decomposition by environmental conditions, including soil moisture and temperature. However, contemporary analytical approaches suggest that the chemical composition of these pools is not necessarily predictable because, despite considerable progress with understanding decomposition processes and the role of decomposer organisms, along with refinements in simulation models, little progress has been made in reconciling biochemical properties with the kinetically defined pools. In this review, we will explore how advances in quantitative analytical techniques have redefined the new understanding of SOM dynamics and how this is affecting the development and application of new modelling approaches to soil C.     

Assessing labile organic carbon in soils undergoing land use change in Brazil: A comparison of approaches

The labile organic C (LC) and C management index (CMI) may be useful indicators of alterations of soil organic matter (SOM) in areas undergoing land use change (LUC) for biofuels production. However, there is no consensus on which methodology is best suited for quantifying LC and CMI. The main goal of this study was to assess alterations on LC contents and CMI values in sites undergoing the LUC native vegetation (NV)-pasture (PA)-sugarcane (SC) in south-central Brazil, and evaluated the sensitivity of different methods commonly used to assess LC and CMI, in order to select a best-suited method to quantify these indicators in tropical regions. The conversion NV-PA decreased the LC and CMI, whilst the conversion of PA-SC tended to increase the CMI. Accordingly, cropping sugarcane in areas previously used as pastures, as currently has been observed in Brazil, enhance SOM quality. The methodology used to quantify the LC and the CMI is critical to infer about LUC effects. Both methods proposed by Blair et al. (1995) and Diekow et al. (2005) were highly sensitive to the conversions evaluated in this research. However, Diekow et al. (2005) is the most suitable method to estimate the LC and CMI in sites undergoing LUC in Brazil, since the approach of Blair et al. (1995) notably overestimates these SOM quality indicators. We reiterate that the SOM changes are well expressed by the total soil organic C in areas undergoing LUC and, integrated approaches, such as the CMI, are quite suitable to evaluate the effects of LUC on SOM.     

Optimization of Biomass and Compost Management to Sustain Soil Organic Matter in Energy Cane Cropping Systems in a Tropical Polluted Soil: a Modelling Study

In French West Indies, the high dependency of the electricity mix on imported fossil fuels has led local authorities to propose the conversion of some land to the production of energy cane. This conversion mainly concerns land polluted by the pesticide chlordecone, where most crops for human consumption have been banned. This molecule has a strong affinity for soil organic matter (SOM). The aims of this study were to assess the impact of crop residue management and compost application on the stocks of SOM and chlordecone in soils cultivated with energy cane and to determine the minimum SOM input required to maintain SOM stocks. A field experiment was conducted to determine the yield and biomass partitioning of energy cane, and laboratory incubations were performed to estimate humification from crop residues. Changes in SOM and chlordecone stocks over a 30-year period were investigated using models already calibrated for the land under study. Non-harvestable biomass left on the field (tops, litterfall and roots) covered >60 % of SOM mineralization. A full offset of mineralization required the return of 10 % of harvestable biomass or the application of compost at a rate of 40 Mg ha^?1 every 5 years. With the total removal of harvestable biomass and without compost applications, SOM and chlordecone losses increased by 23 and 13 %, respectively, which was associated with high SOM mineralization and chlordecone leaching under tropical climate. The estimated break-even price for cane biomass indicated that compost application would be more profitable for farmers than the return of a part of the harvestable biomass.     

Mapping soil organic matter using the topographic wetness index: A comparative study based on different flow-direction algorithms and kriging methods

Terrain attributes derived from digital elevation models have been used widely for mapping soil organic matter (SOM). Among these attributes, the topographic wetness index (TWI), an index for quantitatively indicating the balance between water accumulation and drainage conditions at the local scale, has been shown to correlate with SOM. However, TWIs used in most studies are calculated using a single-flow-direction (SFD) algorithm, which assumes that all water from a grid cell flows into only one neighboring cell. This assumption is not always valid, especially in areas with low relief where movement of water may be divergent. To overcome this SFD limitation, a multiple-flow-direction (MFD) algorithm has been developed, which distributes flow from a grid cell to several downslope neighbors. In this study we compared the effect of TWI calculations based on SFD and MFD in predictive mapping of SOM by incorporating them into different kriging methods over a 51.76 km² area in Nenjiang County of northeastern China. We found that the MFD-based TWI was better correlated with SOM than was the SFD-based index. We then compared the accuracies of SOM maps which were derived from MFD-based TWI and SFD-based TWI incorporated by ordinary kriging (OK), simple kriging with varying local means (SKlm), kriging with external drift (KED) and collocated cokriging (CC). The MFD-based TWI, used as a secondary variable in SKlm and CC, outperforms the SFD-based TWI. For the different kriging methods, CC (incorporating either MFD-based TWI or SFD-based TWI) showed the best performance, and OK generated a better result than SKlm and KED. Both the MFD-based TWI and SFD-based TWI proved to be incompatible with KED and SKlm due to their numerical instability caused by the rough TWI surfaces. Among all predictive methods, CC incorporating the MFD-based TWI produced the best results. This is because: (1) the MFD-based TWI is best able to indicate quantitatively soil moisture and therefore has the strongest correlation with SOM; (2) CC is capable of utilizing effectively the spatial auto-correlation of SOM and the cross-correlation between SOM and the MFD-based TWI.     

The Informational Concept of Searching for Periodicity in Symbol Sequences

A method of informational decomposition has been developed, allowing one to reveal hidden periodicity in any symbol sequence. The informational decomposition is calculated without conversion of a symbol sequence into a numerical one, which facilitates finding periodicities in a symbol sequence. The method permits introducing an analog of the autocorrelation function of a symbol sequence. The method developed by us has been applied to reveal hidden periodicities in nucleotide and amino acid sequences, as well as in different poetical texts. Hidden periodicity has been detected in various genes, testifying to their quantum structure. The functional and structural role of hidden periodicity is discussed.