Combining meta-learning and search techniques to select parameters for support vector machines

Support Vector Machines (SVMs) have achieved very good performance on different learning problems. However, the success of SVMs depends on the adequate choice of the values of a number of parameters (e.g., the kernel and regularization parameters). In the current work, we propose the combination of meta-learning and search algorithms to deal with the problem of SVM parameter selection. In this combination, given a new problem to be solved, meta-learning is employed to recommend SVM parameter values based on parameter configurations that have been successfully adopted in previous similar problems. The parameter values returned by meta-learning are then used as initial search points by a search technique, which will further explore the parameter space. In this proposal, we envisioned that the initial solutions provided by meta-learning are located in good regions of the search space (i.e. they are closer to optimum solutions). Hence, the search algorithm would need to evaluate a lower number of candidate solutions when looking for an adequate solution. In this work, we investigate the combination of meta-learning with two search algorithms: Particle Swarm Optimization and Tabu Search. The implemented hybrid algorithms were used to select the values of two SVM parameters in the regression domain. These combinations were compared with the use of the search algorithms without meta-learning. The experimental results on a set of 40 regression problems showed that, on average, the proposed hybrid methods obtained lower error rates when compared to their components applied in isolation.

Isolation and characterization of sodium 2-fluoroacetate from Mascagnia rigida using chromatography and infrared spectroscopy

Sodium monofluoroacetate was first identified in Dichapetalum cymosum, a South African plant that can cause livestock poisoning and death. After, several other plants also showed to contain this toxin, which leads to the "sudden death". Mascagnia rigida, a well identified poisonous plant, commonly found in northeast of Brazil also cause sudden death in cattle, which shows clinical signs similar to those produced by the ingestion of plants that contain monofluoroacetate. Our aim was to identify the toxic compound present in the aqueous extract of M. rigida. For this purpose, the dried and milled plant was extracted; the extract was lyophilized and submitted to successive chromatographic process, until the desired purity of the active compound was achieved. The study of this material by planar chromatography and by infrared spectrometry indicated that the toxin can be a mixture of mono, di and trifluoroacetate. (C) 2012 Elsevier Ltd. All rights reserved.

Semi-supervised dimensionality reduction based on partial least squares for visual analysis of high dimensional data

Dimensionality reduction is employed for visual data analysis as a way to obtaining reduced spaces for high dimensional data or to mapping data directly into 2D or 3D spaces. Although techniques have evolved to improve data segregation on reduced or visual spaces, they have limited capabilities for adjusting the results according to user's knowledge. In this paper, we propose a novel approach to handling both dimensionality reduction and visualization of high dimensional data, taking into account user's input. It employs Partial Least Squares (PLS), a statistical tool to perform retrieval of latent spaces focusing on the discriminability of the data. The method employs a training set for building a highly precise model that can then be applied to a much larger data set very effectively. The reduced data set can be exhibited using various existing visualization techniques. The training data is important to code user's knowledge into the loop. However, this work also devises a strategy for calculating PLS reduced spaces when no training data is available. The approach produces increasingly precise visual mappings as the user feeds back his or her knowledge and is capable of working with small and unbalanced training sets.

Effect of dental tissue conditioners and matrix metalloproteinase inhibitors on type I collagen microstructure analyzed by Fourier transform infrared spectroscopy

This study aimed to evaluate the chemical interaction of collagen with some substances usually applied in dental treatments to increase the durability of adhesive restorations to dentin. Initially, the similarity between human dentin collagen and type I collagen obtained from commercial bovine membranes of Achilles deep tendon was compared by the Attenuated Total Reflectance technique of Fourier Transform Infrared (ATR-FTIR) spectroscopy. Finally, the effects of application of 35% phosphoric acid, 0.1M ethylenediaminetetraacetic acid (EDTA), 2% chlorhexidine, and 6.5% proanthocyanidin solution on microstructure of collagen and in the integrity of its triple helix were also evaluated by ATR-FTIR. It was observed that the commercial type I collagen can be used as an efficient substitute for demineralized human dentin in studies that use spectroscopy analysis. The 35% phosphoric acid significantly altered the organic content of amides, proline and hydroxyproline of type I collagen. The surface treatment with 0.1M EDTA, 2% chlorhexidine, or 6.5% proanthocyanidin did not promote deleterious structural changes to the collagen triple helix. The application of 6.5% proanthocyanidin on collagen promoted hydrogen bond formation. (c) 2012 Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater, 2012.

Can neuroimaging be used as a support to diagnosis of borderline personality disorder? An approach based on computational neuroanatomy and machine learning

Several recent studies in literature have identified brain morphological alterations associated to Borderline Personality Disorder (BPD) patients. These findings are reported by studies based on voxel-based-morphometry analysis of structural MRI data, comparing mean gray-matter concentration between groups of BPD patients and healthy controls. On the other hand, mean differences between groups are not informative about the discriminative value of neuroimaging data to predict the group of individual subjects. In this paper, we go beyond mean differences analyses, and explore to what extent individual BPD patients can be differentiated from controls (25 subjects in each group), using a combination of automated-morphometric tools for regional cortical thickness/volumetric estimation and Support Vector Machine classifier. The approach included a feature selection step in order to identify the regions containing most discriminative information. The accuracy of this classifier was evaluated using the leave-one-subject-out procedure. The brain regions indicated as containing relevant information to discriminate groups were the orbitofrontal, rostral anterior cingulate, posterior cingulate, middle temporal cortices, among others. These areas, which are distinctively involved in emotional and affect regulation of BPD patients, were the most informative regions to achieve both sensitivity and specificity values of 80% in SVM classification. The findings suggest that this new methodology can add clinical and potential diagnostic value to neuroimaging of psychiatric disorders. (C) 2012 Elsevier Ltd. All rights reserved.

Network-based high level data classification

Traditional supervised data classification considers only physical features (e. g., distance or similarity) of the input data. Here, this type of learning is called low level classification. On the other hand, the human (animal) brain performs both low and high orders of learning and it has facility in identifying patterns according to the semantic meaning of the input data. Data classification that considers not only physical attributes but also the pattern formation is, here, referred to as high level classification. In this paper, we propose a hybrid classification technique that combines both types of learning. The low level term can be implemented by any classification technique, while the high level term is realized by the extraction of features of the underlying network constructed from the input data. Thus, the former classifies the test instances by their physical features or class topologies, while the latter measures the compliance of the test instances to the pattern formation of the data. Our study shows that the proposed technique not only can realize classification according to the pattern formation, but also is able to improve the performance of traditional classification techniques. Furthermore, as the class configuration's complexity increases, such as the mixture among different classes, a larger portion of the high level term is required to get correct classification. This feature confirms that the high level classification has a special importance in complex situations of classification. Finally, we show how the proposed technique can be employed in a real-world application, where it is capable of identifying variations and distortions of handwritten digit images. As a result, it supplies an improvement in the overall pattern recognition rate.

EFFICIENT FAULT LOCATION IN UNDERGROUND DISTRIBUTION SYSTEMS THROUGH OPTIMUM-PATH FOREST

In this article we propose an efficient and accurate method for fault location in underground distribution systems by means of an Optimum-Path Forest (OPF) classifier. We applied the time domains reflectometry method for signal acquisition, which was further analyzed by OPF and several other well-known pattern recognition techniques. The results indicated that OPF and support vector machines outperformed artificial neural networks and a Bayesian classifier, but OPF was much more efficient than all classifiers for training, and the second fastest for classification.

Direct Determination of Sugar Cane Quality Parameters by X-ray Spectrometry and Multivariate Analysis

Current methods for quality control of sugar cane are performed in extracted juice using several methodologies, often requiring appreciable time and chemicals (eventually toxic), making the methods not green and expensive. The present study proposes the use of X-ray spectrometry together with chemometric methods as an innovative and alternative technique for determining sugar cane quality parameters, specifically sucrose concentration, POL, and fiber content. Measurements in stem, leaf, and juice were performed, and those applied directly in stem provided the best results. Prediction models for sugar cane stem determinations with a single 60 s irradiation using portable X-ray fluorescence equipment allows estimating the % sucrose, % fiber, and POL simultaneously. Average relative deviations in the prediction step of around 8% are acceptable if considering that field measurements were done. These results may indicate the best period to cut a particular crop as well as for evaluating the quality of sugar cane for the sugar and alcohol industries.

Conformational preferences for some 3-(4 '-substituted phenylsulfonyl)-1-methyl-2-piperidones through spectroscopic and theoretical studies

The analysis of the infrared (IR) carbonyl band of some 3-(4'-substituted phenylsulfonyl)-1-methyl-2-piperidones 1-5 bearing as substituents: OMe 1, Me 2, H 3, Cl 4 and NO2 5, supported by B3LY13/6-31G(d,p) calculations along with NBO analysis (for 1, 3 and 5) and X-ray diffraction (for 5), indicated the existence of three stable conformations i.e. quasi-axial (q-ax), syn-clinal (s-cl) and quasi-equatorial (q-eq). In the gas phase, the q-ax conformer is calculated as the most stable (ca. 88%) and the least polar, the s-cl conformer is less stable (ca. 12%) but more polar, and the q-eq conformer is the least stable (ca. 1%) and the most polar of the three conformers evaluated. The sum of the most important orbital interactions from NBO analysis and the trend of the electrostatic interactions accounts for the relative populations as well as for the v(CO) frequencies of the q-ax. s-cl and q-eq conformers calculated in the gas phase. The unique IR v(CO) band in CCl4 may be ascribed to the most stable q-ax conformer. The more intense (60%) high frequency doublet component in CHCl3 may be assigned to the summing up of the least stable q-eq and the less stable s-cl conformers, as their frequencies are almost coincident. The occurrence of only a single v(CO) band in both CH2Cl2 and CH3CN supports the fact that the v(CO) band of the two more polar conformers appear as a single band. Additional support to this rationalization is given by the single point PCM method, which showed a progressive increase of the q-eq + s-cl/q-ax population ratio going from the gas phase to CCl4, to CHCl3, to CH2Cl2 and to CN3CN. X-ray single crystal analysis of 5 indicates that this compound displays a quasi-axial geometry with respect to the [O=C-CH-S] moiety, and that the 2-piperidone ring assumes a slightly distorted half-chair conformation. In the crystal packing, molecules of 5 are arranged into supramolecular layers linked through C-H center dot center dot center dot O interactions along with it pi center dot center dot center dot pi interactions between adjacent benzene rings. (C) 2012 Elsevier B.V. All rights reserved.