Effect of oral calcium carbonate on aortic calcification in apolipoprotein E-deficient (apoE-/-) mice with chronic renal failure.

BACKGROUND: In chronic kidney disease (CKD) patients, the intake of calcium-based phosphate binders is associated with a marked progression of coronary artery and aortic calcification, in contrast to patients receiving calcium-free phosphate binders. The aim of this study was to reexamine the role of calcium carbonate in vascular calcification and to analyse its effect on aortic calcification-related gene expression in chronic renal failure (CRF). METHODS: Mice deficient in apolipoprotein E underwent either sham operation or subtotal nephrectomy to create CRF. They were then randomly assigned to one of the three following groups: a control non-CRF group and a CRF group fed on standard diet, and a CRF group fed on calcium carbonate enriched diet, for a period of 8 weeks. Aortic atherosclerotic plaque and calcification were evaluated using quantitative morphologic image processing. Aortic gene and protein expression was examined using immunohistochemistry and Q-PCR methods. RESULTS: Calcium carbonate supplementation was effective in decreasing serum phosphorus but was associated with a higher serum calcium concentration. Compared with standard diet, calcium carbonate enriched diet unexpectedly induced a significant decrease of both plaque (p<0.05) and non-plaque-associated calcification surface (p<0.05) in CRF mice. It also increased osteopontin (OPN) protein expression in atherosclerotic lesion areas of aortic root. There was also a numerical increase in OPN and osteoprotegerin gene expression in total thoracic aorta but the difference did not reach the level of significance. Finally, calcium carbonate did not change the severity of atherosclerotic lesions. CONCLUSION: In this experimental model of CRF, calcium carbonate supplementation did not accelerate but instead decreased vascular calcification. If our observation can be extrapolated to humans, it appears to question the contention that calcium carbonate supplementation, at least when given in moderate amounts, necessarily enhances vascular calcification. It is also compatible with the hypothesis of a preponderant role of phosphorus over that of calcium in promoting vascular calcification in CRF.

Predictability of music descriptor time series and its application to cover song detection

Intuitively, music has both predictable and unpredictable components. In this work we assess this qualitative statement in a quantitative way using common time series models fitted to state-of-the-art music descriptors. These descriptors cover different musical facets and are extracted from a large collection of real audio recordings comprising a variety of musical genres. Our findings show that music descriptor time series exhibit a certain predictability not only for short time intervals, but also for mid-term and relatively long intervals. This fact is observed independently of the descriptor, musical facet and time series model we consider. Moreover, we show that our findings are not only of theoretical relevance but can also have practical impact. To this end we demonstrate that music predictability at relatively long time intervals can be exploited in a real-world application, namely the automatic identification of cover songs (i.e. different renditions or versions of the same musical piece). Importantly, this prediction strategy yields a parameter-free approach for cover song identification that is substantially faster, allows for reduced computational storage and still maintains highly competitive accuracies when compared to state-of-the-art systems.

Morphodynamic analysis of cerebral aneurysm pulsation from time-resolved rotational angiography

This paper presents a technique to estimate and model patient-specific pulsatility of cerebral aneurysms over onecardiac cycle, using 3D rotational X-ray angiography (3DRA) acquisitions. Aneurysm pulsation is modeled as a time varying-spline tensor field representing the deformation applied to a reference volume image, thus producing the instantaneousmorphology at each time point in the cardiac cycle. The estimated deformation is obtained by matching multiple simulated projections of the deforming volume to their corresponding original projections. A weighting scheme is introduced to account for the relevance of each original projection for the selected time point. The wide coverage of the projections, together with the weighting scheme, ensures motion consistency in all directions. The technique has been tested on digital and physical phantoms that are realistic and clinically relevant in terms of geometry, pulsation and imaging conditions. Results from digital phantomexperiments demonstrate that the proposed technique is able to recover subvoxel pulsation with an error lower than 10% of the maximum pulsation in most cases. The experiments with the physical phantom allowed demonstrating the feasibility of pulsation estimation as well as identifying different pulsation regions under clinical conditions.

Flow-R, a model for susceptibility mapping of debris flows and other gravitational hazards at a regional scale

The development of susceptibility maps for debris flows is of primary importance due to population pressure in hazardous zones. However, hazard assessment by processbased modelling at a regional scale is difficult due to the complex nature of the phenomenon, the variability of local controlling factors, and the uncertainty in modelling parameters. A regional assessment must consider a simplified approach that is not highly parameter dependant and that can provide zonation with minimum data requirements. A distributed empirical model has thus been developed for regional susceptibility assessments using essentially a digital elevation model (DEM). The model is called Flow-R for Flow path assessment of gravitational hazards at a Regional scale (available free of charge under www.flow-r.org) and has been successfully applied to different case studies in various countries with variable data quality. It provides a substantial basis for a preliminary susceptibility assessment at a regional scale. The model was also found relevant to assess other natural hazards such as rockfall, snow avalanches and floods. The model allows for automatic source area delineation, given user criteria, and for the assessment of the propagation extent based on various spreading algorithms and simple frictional laws.We developed a new spreading algorithm, an improved version of Holmgren's direction algorithm, that is less sensitive to small variations of the DEM and that is avoiding over-channelization, and so produces more realistic extents. The choices of the datasets and the algorithms are open to the user, which makes it compliant for various applications and dataset availability. Amongst the possible datasets, the DEM is the only one that is really needed for both the source area delineation and the propagation assessment; its quality is of major importance for the results accuracy. We consider a 10m DEM resolution as a good compromise between processing time and quality of results. However, valuable results have still been obtained on the basis of lower quality DEMs with 25m resolution.

Transcriptional and functional profiles of voltage-gated Na(+) channels in injured and non-injured DRG neurons in the SNI model of neuropathic pain.

Changes in expression and function of voltage-gated sodium channels (VGSC) in dorsal root ganglion (DRG) neurons may play a major role in the genesis of peripheral hyperexcitability that occurs in neuropathic pain. We present here the first description of changes induced by spared nerve injury (SNI) to Na(v)1 mRNA levels and tetrodotoxin-sensitive and -resistant (TTX-S/TTX-R) Na(+) currents in injured and adjacent non-injured small DRG neurons. VGSC transcripts were down-regulated in injured neurons except for Na(v)1.3, which increased, while they were either unchanged or increased in non-injured neurons. TTX-R current densities were reduced in injured neurons and the voltage dependence of steady-state inactivation for TTX-R was positively shifted in injured and non-injured neurons. TTX-S current densities were not affected by SNI, while the rate of recovery from inactivation was accelerated in injured neurons. Our results describe altered neuronal electrogenesis following SNI that is likely induced by a complex regulation of VGSCs.

The role of Bmi1 in CNS development and in retinal degeneration

SUMMARY : The present work addresses several aspects of cell cycle regulation, cell fate specification and cell death in the central nervous system (CNS), specifically the cortex and the retina. More precisely, we investigated the role of Bmi1, a polycomb family gene required for stem cell proliferation and self-renewal, in the development of the cerebral cortex, as well as in the genesis of the retina. These data, together with studies published during the last two decades concerning cell cycle re-activation in apoptotic neurons in the CNS, raised the question of a possible link between regulation of the cell cycle during development and during retinal degeneration. 1. The effects of Bmi1 loss in the cerebral cortex : Consistently with our and others' observations on failure of Bmi9-/- stem cells to proliferate and self-renew in vitro, the Bmi9-/- cerebral cortex presented slight defects in proliferation in stem/progenitor cells compartments in vivo. This was in accordance with the pattern of Bmi1 expression in the developing forebrain. The modest proliferation defects, compared to the drastic consequences of Bmi9 loss in vitro, suggest that cell-extrinsic mechanisms may partially compensate for Bmi1 deletion in vivo during cortical histogenesis. Nevertheless, we observed a decreased proliferating activity in neurogenic regions of the adult telencephalon, more precisely in the subventricular zone, showing that Bmi1 controls neural stem/progenitor proliferation during adulthood in vivo. Our data also highlight an increased production of astrocytes at birth, and a generalized gliosis in the adult Bmi9-/- brain. Importantly, glial progenitors and astrocytes retained the ability to proliferate in the absence of Bmi1. 2. The effects of Bmi1 loss in the retina : The pattern of expression of Bmi1 during development and in the adult retina suggests a role for Bmi1 in cell fate specification and differentiation rather than in proliferation. While the layering and the global structure of the retina appear normal in Bmi1 /adult mice, immunohistochemìcal analysis revealed defects in the three major classes of retinal interneurons, namely: horizontal, bipolar and amacrine cells. Electroretinogram recordings in Bmi9-/- mice are coherent with the defects observed at the histological level, with a reduced b-wave and low-profile oscillatory potentials. These results show that Bmi1 controls not only proliferation, but also cell type generation, as previously observed in the cerebellum. 3. Cell cycle events and related neuroprotective strategies in retinal degeneration : In several neurodegenerative disorders, neurons re-express cell cycle proteins such as cyclin dependent kinases (Cdks) prior to apoptosis. Here, we show for the first time that this is also the case during retinal degeneration. Rd1 mice carry a recessive defect (Pdeóbrd/rd) that causes retinal degeneration and serves as a model of retinitis pigmentosa. We found that photoreceptors express Cdk4 and Cdk2, and undergo DNA synthesis prior to cell death. To interfere with the reactivation of Cdk-related pathways, we deleted E2fs or Brni1, which normally allow cell cycle progression. While deleting E2f1 (downstream of Cdk4/6) in Rd1 mice provides only temporary protection, knocking out Bmi1 (upstream of Cdks) leads to an extensive neuroprotective effect, independent of p16ink4a or p19arf, two tumor suppressors regulated by Bmi1. Analysis of Cdks and the DNA repair-related protein Ligase IV showed that Bmi1 acts downstream of DNA repair events and upstream of Cdks in this neurodegenerative mechanism. Expression of Cdks during an acute model of retinal degeneration, light damage-induced photoreceptor death, points to a role for Bmi1 and cell cycle proteins in retinal degeneration. Considering the similarity with the cell cycle-related apoptotic pathway observed in other neurodegenerative diseases, Bmi1 is a possible general target to prevent or delay neuronal death. RESUME : Ce travail aborde plusieurs aspects de la régulation du cycle cellulaire, de la spécification du devenir des cellules et de la mort cellulaire dans le système nerveux centrale (SNC), plus particulièrement dans le cortex cérébral et dans la rétine. Nous nous sommes intéressés au gène Bmi1, appartenant à la famille polycomb et nécessaire à la prolifération et au renouvellement des cellules souches. Nous avons visé à disséquer son rôle dans le développement du cortex et de la rétine. Ces données, ainsi qu'une série de travaux publiés au cours des deux dernières décennies concernant la réactivation du cycle cellulaire dans les neurones en voie d'apoptose dans le SNC, nous ont ensuite poussé à chercher un lien entre la régulation du cycle cellulaire pendant le développement et au cours de la dégénérescence rétinienne. 1. Les effets de l'inactivation de Bmi1 dans le cortex cérébral : En accord avec l'incapacité des cellules souches neurales in vitro à proliférer et à se renouveler en absence de Bmi1, le cortex cérébral des souris Bmi1-/- présente de légers défauts de prolifération dans les compartiments contenant les cellules souches neurales. Ceci est en accord avec le profil d'expression de Bmi1 dans le télencéphale. Les conséquences de la délétion de Bmi1 sont toutefois nettement moins prononcées in vivo qu'in vitro ; cette différence suggère l'existence de mécanismes pouvant partiellement compenser l'absence de Bmi1 pendant la corticogenèse. Néanmoins, l'observation d'une réduction de la prolifération dans la zone sous-ventriculaire, la zone majeure de neurogenèse dans le télencéphale adulte, montre que Bmi1 contrôle la prolifération des cellules souche/progénitrices neurales chez la souris adulte. Nos résultats démontrent par ailleurs une augmentation de la production d'astrocytes à la naissance ainsi qu'une gliose généralisée à l'état adulte chez les souris Bmi1-/-. Les progéniteurs gliaux et les astrocytes conservent donc leur capacité à proliférer en absence de Bmi1. 2. Les effets de l'inactivation de Bmi1 dans la rétine : Le profil d'expression de Bmi1 pendant fe développement ainsi que dans la rétine adulte suggère un rôle de Bmi1 dans la spécification de certains types cellulaires et dans la différentiation plutôt que dans la prolifération. Alors que la structure et la lamination de la rétine semblent normales chez les souris Bmi1-/-, l'analyse par immunohistochimie amis en évidence des défauts au niveau des trois classes d'interneurones rétiniens (les cellules horizontales, bipolaires et amacrines). Les électrorétinogrammes des souris Bmi1-/- sont cohérents avec les défauts observés au niveau histologique et montrent une réduction de l'onde « b » et des potentiels oscillatoires. Ces résultats montrent que Bmi1 contrôle la génération de certaines sous-populations de neurones, comme démontré auparavant au niveau de cervelet. 3. Réactivation du cycle cellulaire et stratégies théraoeutiaues dans les dégénérescences rétiniennes : Dans plusieurs maladies neurodégénératives, les neurones ré-expriment des protéines du cycle cellulaire telles que les kinases cycline-dépendantes (Cdk) avant d'entrer en apoptose. Nous avons démontré que c'est aussi le cas dans les dégénérescences rétiniennes. Les souris Rd1 portent une mutation récessive (Pde6brd/rd) qui induit une dégénérescence de la rétine et sont utilisées comme modèle animal de rétinite pigmentaire. Nous avons observé que les photorécepteurs expriment Cdk4 et Cdk2, et entament une synthèse d'ADN avant de mourir par apoptose. Pour interférer avec la réactivation les mécanismes Cdk-dépendants, nous avons inactivé les gènes E2f et Bmi1, qui permettent normalement la progression du cycle cellulaire. Nous avons mis en évidence que la délétion de E2f1 (en aval de Cdk4/6) dans les souris Rd1 permet une protection transitoire des photorécepteurs. Toutefois, l'inactivation de Bmi1 (en amont des Cdk) est corrélée à une neuroprotection bien plus durable et ceci indépendamment de p16ink4a et p19arf, deux suppresseurs de tumeurs normalement régulés par Bmi1. L'analyse des Cdk et de la ligase IV (une protéine impliquée dans les mécanismes de réparation de l'ADN) a montré que Bmi1 agit en aval des événements de réparation de l'ADN et en amont des Cdk dans la cascade apoptotique dans les photorécepteurs des souris Rd1. Nous avons également observé la présence de Cdk dans un modèle aigu de dégénérescence rétinienne induit par une exposition des animaux à des niveaux toxiques de lumière. Nos résultats suggèrent donc un rôle général de Bmi1 et des protéines du cycle cellulaire dans les dégénérescences de la rétine. Si l'on considère la similarité avec les événements de réactivation du cycle cellulaire observés dans d'autres maladies neurodégénératives, Bmi1 pourrait être une cible thérapeutique générale pour prévenir la mort neuronale.

Randomized observation periods for the compound Poisson risk model: Dividends

In the framework of the classical compound Poisson process in collective risk theory, we study a modification of the horizontal dividend barrier strategy by introducing random observation times at which dividends can be paid and ruin can be observed. This model contains both the continuous-time and the discrete-time risk model as a limit and represents a certain type of bridge between them which still enables the explicit calculation of moments of total discounted dividend payments until ruin. Numerical illustrations for several sets of parameters are given and the effect of random observation times on the performance of the dividend strategy is studied.

Profile order and time-dependent artifacts in contrast-enhanced coronary MR angiography at 3T: origin and prevention.

To enhance the clinical value of coronary magnetic resonance angiography (MRA), high-relaxivity contrast agents have recently been used at 3T. Here we examine a uniform bilateral shadowing artifact observed along the coronary arteries in MRA images collected using such a contrast agent. Simulations were performed to characterize this artifact, including its origin, to determine how best to mitigate this effect, and to optimize a data acquisition/injection scheme. An intraluminal contrast agent concentration model was used to simulate various acquisition strategies with two profile orders for a slow-infusion of a high-relaxivity contrast agent. Filtering effects from temporally variable weighting in k-space are prominent when a centric, radial (CR) profile order is applied during contrast infusion, resulting in decreased signal enhancement and underestimation of vessel width, while both pre- and postinfusion steady-state acquisitions result in overestimation of the vessel width. Acquisition during the brief postinfusion steady-state produces the greatest signal enhancement and minimizes k-space filtering artifacts.

The influence of Bone Morphogenic Protein-2 on the consolidation phase in a distraction osteogenesis model.

We asked whether locally applied recombinant-Bone Morphogenic Protein-2 (rh-BMP-2) with an absorbable Type I collagen sponge (ACS) carrier could enhance the consolidation phase in a callotasis model. We performed unilateral transverse osteotomy of the tibia in 21 immature male rabbits. After a latency period of 7 days, a 3-weeks distraction was begun at a rate of 0.5mm/12h. At the end of the distraction period (Day 28) animals were randomly divided into three groups and underwent a second surgical procedure: 6 rabbits in Group I (Control group; the callus was exposed and nothing was added), 6 rabbits in Group II (ACS group; receiving the absorbable collagen sponge soaked with saline) and 9 rabbits in Group III (rh-BMP-2/ACS group; receiving the ACS soaked with 100μg/kg of rh-BMP-2, Inductos(®), Medtronic). Starting at Day 28 we assessed quantitative and qualitative radiographic parameters as well as densitometric parameters every two weeks (Days 28, 42, 56, 70 and 84). Animals were sacrificed after 8 weeks of consolidation (Day 84). Qualitative radiographic evaluation revealed hypertrophic calluses in the Group III animals. The rh-BMP-2/ACS also influenced the development of the cortex of the calluses as shown by the modified radiographic patterns in Group III when compared to Groups I and II. Densitometric analysis revealed the bone mineral content (BMC) was significantly higher in the rh-BMP-2/ACS treated animals (Group III).

A developmental in vivo 1H NMR study in mice with genetic redox dysregulation: an animal model with relevance to schizophrenia

Glutathione (GSH), a major redox regulator and anti-oxidant, is decreased in cerebrospinal fluid and prefrontal cortex of schizophrenia patients. The gene of the key GSH-synthesizing enzyme, glutamate-cysteine ligase, modifier (GCLM) subunit, is associated with schizophrenia, suggesting that the deficit in the GSH system is of genetic origin. Using the GCLM knock-out (KO) mouse model with 60% decreased brain GSH levels, we have shown that redox dysregulation results in abnormal brain morphology and function. Current theory holds that schizophrenia is a developmental disease involving progressive anatomical and functional brain pathology. Here, we used GCLM KO mice to investigate the impact of a genetically dysregulated redox system on the neurochemical profile of the developing brain. The anterior and posterior cortical neurochemical profile of male and female GCLM KO, heterozygous and wildtype mice was determined by localised in vivo 1H NMR spectroscopy at 14.1 T (Varian/Magnex spectrometer) on post-natal days 10, 20, 30, 60 and 90. We show, for the first time, (1) that high quality 1H NMR spectra can be acquired from early developing mouse brains and (2) that recurrent anaesthesia by itself when administered at the same developmental days has no adverse effects on brain metabolites nor on adult behaviour. (3) Most importantly, our results reveal genotype and age specific changes for a number of metabolites revealing insight into normal brain development and about the impact of genetic GSH dysregulation.

Pivotal role for a tail subunit of the RNA polymerase II mediator complex CgMed2 in azole tolerance and adherence in Candida glabrata.

Antifungal therapy failure can be associated with increased resistance to the employed antifungal agents. Candida glabrata, the second most common cause of invasive candidiasis, is intrinsically less susceptible to the azole class of antifungals and accounts for 15% of all Candida bloodstream infections. Here, we show that C. glabrata MED2 (CgMED2), which codes for a tail subunit of the RNA polymerase II Mediator complex, is required for resistance to azole antifungal drugs in C. glabrata. An inability to transcriptionally activate genes encoding a zinc finger transcriptional factor, CgPdr1, and multidrug efflux pump, CgCdr1, primarily contributes to the elevated susceptibility of the Cgmed2Δ mutant toward azole antifungals. We also report for the first time that the Cgmed2Δ mutant exhibits sensitivity to caspofungin, a constitutively activated protein kinase C-mediated cell wall integrity pathway, and elevated adherence to epithelial cells. The increased adherence of the Cgmed2Δ mutant was attributed to the elevated expression of the EPA1 and EPA7 genes. Further, our data demonstrate that CgMED2 is required for intracellular proliferation in human macrophages and modulates survival in a murine model of disseminated candidiasis. Lastly, we show an essential requirement for CgMed2, along with the Mediator middle subunit CgNut1 and the Mediator cyclin-dependent kinase/cyclin subunit CgSrb8, for the high-level fluconazole resistance conferred by the hyperactive allele of CgPdr1. Together, our findings underscore a pivotal role for CgMed2 in basal tolerance and acquired resistance to azole antifungals.

Non-Quasi-Competitiveness and Stability in Cournot's Model

This paper retakes previous work of the authors, about the relationship between non-quasi-competitiveness (the increase in price caused by an increase in the number of oligopolists) and stability of the equilibrium in the classical Cournot oligopoly model. Though it has been widely accepted in the literature that the loss of quasi-competitiveness is linked, in the long run as new firms entered the market, to instability of the model, the authors in their previous work put forward a model in which a situation of monopoly changed to duopoly losing quasi-competitiveness but maintaining the stability of the equilibrium. That model could not, at the time, be extended to any number of oligopolists. The present paper exhibits such an extension. An oligopoly model is shown in which the loss of quasi-competitiveness resists the presence in the market of as many firms as one wishes and where the successive Cournot's equilibrium points are unique and asymptotically stable. In this way, for the first time, the conjecture that non-quasi- competitiveness and instability were equivalent in the long run, is proved false.

Focused time-lapse inversion of radio and audio magnetotelluric data

Geoelectrical techniques are widely used to monitor groundwater processes, while surprisingly few studies have considered audio (AMT) and radio (RMT) magnetotellurics for such purposes. In this numerical investigation, we analyze to what extent inversion results based on AMT and RMT monitoring data can be improved by (1) time-lapse difference inversion; (2) incorporation of statistical information about the expected model update (i.e., the model regularization is based on a geostatistical model); (3) using alternative model norms to quantify temporal changes (i.e., approximations of l(1) and Cauchy norms using iteratively reweighted least-squares), (4) constraining model updates to predefined ranges (i.e., using Lagrange Multipliers to only allow either increases or decreases of electrical resistivity with respect to background conditions). To do so, we consider a simple illustrative model and a more realistic test case related to seawater intrusion. The results are encouraging and show significant improvements when using time-lapse difference inversion with non l(2) model norms. Artifacts that may arise when imposing compactness of regions with temporal changes can be suppressed through inequality constraints to yield models without oscillations outside the true region of temporal changes. Based on these results, we recommend approximate l(1)-norm solutions as they can resolve both sharp and smooth interfaces within the same model. (C) 2012 Elsevier B.V. All rights reserved.

Hormonal, global, and regional haemodynamic responses to a vascular antagonist of vasopressin in patients with congestive heart failure with and without hyponatraemia.

The pathophysiological role of an increase in circulating vasopressin in sustaining global and regional vasoconstriction in patients with congestive heart failure has not been established, particularly in patients with hyponatraemia. To assess this further, 20 patients with congestive heart failure refractory to digoxin and diuretics were studied before and 60 minutes after the intravenous injection (5 micrograms/kg) of the vascular antagonist of vasopressin [1(beta-mercapto-beta,beta-cyclopentamethylene-propionic acid), 2-(0-methyl) tyrosine] arginine vasopressin. Ten patients were hyponatraemic (plasma sodium less than 135 mmol/l) and 10 were normonatraemic. In both groups of patients the vascular vasopressin antagonist did not alter systemic or pulmonary artery pressures, right atrial pressure, pulmonary capillary wedge pressure, cardiac index, or vascular resistances. Furthermore, there was no change in skin and hepatic blood flow in either group after the injection of the vascular antagonist. Only one patient in the hyponatraemic group showed considerable haemodynamic improvement. He had severe congestive heart failure and a high concentration of plasma vasopressin (51 pmol/l). Plasma renin activity, vasopressin, or catecholamine concentrations were not significantly changed in response to the administration of the vasopressin antagonist in either the hyponatraemic or the normonatraemic groups. Patients with hyponatraemia, however, had higher baseline plasma catecholamine concentrations, heart rate, pulmonary pressure and resistance, and lower hepatic blood flow than patients without hyponatraemia. Plasma vasopressin and plasma renin activity were slightly, though not significantly, higher in the hyponatraemic group. Thus the role of vasopressin in sustaining regional or global vasoconstriction seems limited in patients with congestive heart failure whether or not concomitant hyponatraemia is present. Vasopressin significantly increases the vascular tone only in rare patients with severe congestive heart failure and considerably increased vasopressin concentrations. Patients with hyponatraemia do, however, have raised baseline catecholamine concentrations, heart rate, pulmonary arterial pressure and resistance, and decreased hepatic blood flow.

Time Series Input Selection using Multiple Kernel Learning

In this paper we study the relevance of multiple kernel learning (MKL) for the automatic selection of time series inputs. Recently, MKL has gained great attention in the machine learning community due to its flexibility in modelling complex patterns and performing feature selection. In general, MKL constructs the kernel as a weighted linear combination of basis kernels, exploiting different sources of information. An efficient algorithm wrapping a Support Vector Regression model for optimizing the MKL weights, named SimpleMKL, is used for the analysis. In this sense, MKL performs feature selection by discarding inputs/kernels with low or null weights. The approach proposed is tested with simulated linear and nonlinear time series (AutoRegressive, Henon and Lorenz series).