Monitoring network optimisation for spatial data classification using support vector machines

The paper presents a novel method for monitoring network optimisation, based on a recent machine learning technique known as support vector machine. It is problem-oriented in the sense that it directly answers the question of whether the advised spatial location is important for the classification model. The method can be used to increase the accuracy of classification models by taking a small number of additional measurements. Traditionally, network optimisation is performed by means of the analysis of the kriging variances. The comparison of the method with the traditional approach is presented on a real case study with climate data.

Multiple Kernel Learning of Environmental Data. Case study: Analysis and Mapping of Wind Fields

The paper presents the Multiple Kernel Learning (MKL) approach as a modelling and data exploratory tool and applies it to the problem of wind speed mapping. Support Vector Regression (SVR) is used to predict spatial variations of the mean wind speed from terrain features (slopes, terrain curvature, directional derivatives) generated at different spatial scales. Multiple Kernel Learning is applied to learn kernels for individual features and thematic feature subsets, both in the context of feature selection and optimal parameters determination. An empirical study on real-life data confirms the usefulness of MKL as a tool that enhances the interpretability of data-driven models.

Individual prediction of cognitive decline in mild cognitive impairment using support vector machine-based analysis of diffusion tensor imaging data.

Although cross-sectional diffusion tensor imaging (DTI) studies revealed significant white matter changes in mild cognitive impairment (MCI), the utility of this technique in predicting further cognitive decline is debated. Thirty-five healthy controls (HC) and 67 MCI subjects with DTI baseline data were neuropsychologically assessed at one year. Among them, there were 40 stable (sMCI; 9 single domain amnestic, 7 single domain frontal, 24 multiple domain) and 27 were progressive (pMCI; 7 single domain amnestic, 4 single domain frontal, 16 multiple domain). Fractional anisotropy (FA) and longitudinal, radial, and mean diffusivity were measured using Tract-Based Spatial Statistics. Statistics included group comparisons and individual classification of MCI cases using support vector machines (SVM). FA was significantly higher in HC compared to MCI in a distributed network including the ventral part of the corpus callosum, right temporal and frontal pathways. There were no significant group-level differences between sMCI versus pMCI or between MCI subtypes after correction for multiple comparisons. However, SVM analysis allowed for an individual classification with accuracies up to 91.4% (HC versus MCI) and 98.4% (sMCI versus pMCI). When considering the MCI subgroups separately, the minimum SVM classification accuracy for stable versus progressive cognitive decline was 97.5% in the multiple domain MCI group. SVM analysis of DTI data provided highly accurate individual classification of stable versus progressive MCI regardless of MCI subtype, indicating that this method may become an easily applicable tool for early individual detection of MCI subjects evolving to dementia.

Optimal priming of poxvirus vector (NYVAC)-based HIV vaccine regimens requires 3 DNA injections. Results of the EV03/ANRS Vac20 Phase I/II Trial

Background: Two or three DNA primes have been used in previous smaller clinical trials, but the number required for optimal priming of viral vectors has never been assessed in adequately powered clinical trials. The EV03/ANRS Vac20 phase I/II trial investigated this issue using the DNA prime/poxvirus NYVAC boost combination, both expressing a common HIV-1 clade C immunogen consisting of Env and Gag-Pol-Nef polypeptide. Methods: 147 healthy volunteers were randomly allocated through 8 European centres to either 3xDNA plus 1xNYVAC (weeks 0, 4, 8 plus 24; n¼74) or to 2xDNA plus 2xNYVAC (weeks 0, 4 plus 20, 24; n¼73), stratified by geographical region and sex. T cell responses were quantified using the interferon g Elispot assay and 8 peptide pools; samples from weeks 0, 26 and 28 (time points for primary immunogenicity endpoint), 48 and 72 were considered for this analysis. Results: 140 of 147 participants were evaluable at weeks 26 and/ or 28. 64/70 (91%) in the 3xDNA arm compared to 56/70 (80%) in the 2xDNA arm developed a T cell response (P¼0.053). 26 (37%) participants of the 3xDNA arm developed a broader T cell response (Env plus at least to one of the Gag, Pol, Nef peptide pools) versus 15 (22%) in the 2xDNA arm (P¼0.047). At week 26, the overall magnitude of responses was also higher in the 3xDNA than in the 2xDNA arm (similar at week 28), with a median of 545 versus 328 SFUs/106 cells at week 26 (P<0.001). Preliminary overall evaluation showed that participants still developed T-cell response at weeks 48 (78%, n¼67) and 72 (70%, n¼66). Conclusion: This large clinical trial demonstrates that optimal priming of poxvirus-based vaccine regimens requires 3 DNA regimens and further confirms that the DNA/NYVAC prime boost vaccine combination is highly immunogenic and induced durable T-cell responses.

Pituitary blastoma: a pathognomonic feature of germ-line DICER1 mutations.

Individuals harboring germ-line DICER1 mutations are predisposed to a rare cancer syndrome, the DICER1 Syndrome or pleuropulmonary blastoma-familial tumor and dysplasia syndrome [online Mendelian inheritance in man (OMIM) #601200]. In addition, specific somatic mutations in the DICER1 RNase III catalytic domain have been identified in several DICER1-associated tumor types. Pituitary blastoma (PitB) was identified as a distinct entity in 2008, and is a very rare, potentially lethal early childhood tumor of the pituitary gland. Since the discovery by our team of an inherited mutation in DICER1 in a child with PitB in 2011, we have identified 12 additional PitB cases. We aimed to determine the contribution of germ-line and somatic DICER1 mutations to PitB. We hypothesized that PitB is a pathognomonic feature of a germ-line DICER1 mutation and that each PitB will harbor a second somatic mutation in DICER1. Lymphocyte or saliva DNA samples ascertained from ten infants with PitB were screened and nine were found to harbor a heterozygous germ-line DICER1 mutation. We identified additional DICER1 mutations in nine of ten tested PitB tumor samples, eight of which were confirmed to be somatic in origin. Seven of these mutations occurred within the RNase IIIb catalytic domain, a domain essential to the generation of 5p miRNAs from the 5' arm of miRNA-precursors. Germ-line DICER1 mutations are a major contributor to PitB. Second somatic DICER1 "hits" occurring within the RNase IIIb domain also appear to be critical in PitB pathogenesis.

How do the different components of episodic memory develop? Role of executive functions and short-term feature-binding abilities.

This study investigated the development of all 3 components of episodic memory (EM), as defined by Tulving, namely, core factual content, spatial context, and temporal context. To this end, a novel, ecologically valid test was administered to 109 participants aged 4-16 years. Results showed that each EM component develops at a different rate. Ability to memorize factual content emerges early, whereas context retrieval abilities continue to improve until adolescence, due to persistent encoding difficulties (isolated by comparing results on free recall and recognition tasks). Exploration of links with other cognitive functions revealed that short-term feature-binding abilities contribute to all EM components, and executive functions to temporal and spatial context, although ability to memorize temporal context is predicted mainly by age.

Ensemble methods for environmental data modelling with support vector regression

This paper investigates the use of ensemble of predictors in order to improve the performance of spatial prediction methods. Support vector regression (SVR), a popular method from the field of statistical machine learning, is used. Several instances of SVR are combined using different data sampling schemes (bagging and boosting). Bagging shows good performance, and proves to be more computationally efficient than training a single SVR model while reducing error. Boosting, however, does not improve results on this specific problem.

Poxvirus vector-based HIV vaccines.

PURPOSE OF REVIEW: In this review, we will provide the scientific rationale for the use of poxvirus vectors in the field of HIV vaccines, the immunological profile of the vaccine-induced immune responses, an update on the current use of poxvirus vector-based vaccines in HIV vaccine clinical trials, and the development of new modified poxvirus vectors with improved immunological profile. RECENT FINDINGS: An Ad5-HIV vaccine was tested in a phase IIb clinical trial (known as the Step trial). Vaccinations in the Step trial were discontinued because the vaccine did not show any effect on acquisition of infection and on viral load. After the disappointing failure of the Step trial, the field of HIV vaccine has regained enthusiasm and vigour due to the promising protective effect observed in the phase III efficacy trial (known as RV-144) performed in Thailand which has tested a poxvirus-gp120 combination. SUMMARY: The RV-144 phase III has provided for the first time evidence that an HIV vaccine can prevent HIV infection. The results from the RV-144 trial are providing the scientific rationale for the future development of the HIV vaccine field and for designing future efficacy trials.

Data-driven topo-climatic mapping with machine learning methods

Automatic environmental monitoring networks enforced by wireless communication technologies provide large and ever increasing volumes of data nowadays. The use of this information in natural hazard research is an important issue. Particularly useful for risk assessment and decision making are the spatial maps of hazard-related parameters produced from point observations and available auxiliary information. The purpose of this article is to present and explore the appropriate tools to process large amounts of available data and produce predictions at fine spatial scales. These are the algorithms of machine learning, which are aimed at non-parametric robust modelling of non-linear dependencies from empirical data. The computational efficiency of the data-driven methods allows producing the prediction maps in real time which makes them superior to physical models for the operational use in risk assessment and mitigation. Particularly, this situation encounters in spatial prediction of climatic variables (topo-climatic mapping). In complex topographies of the mountainous regions, the meteorological processes are highly influenced by the relief. The article shows how these relations, possibly regionalized and non-linear, can be modelled from data using the information from digital elevation models. The particular illustration of the developed methodology concerns the mapping of temperatures (including the situations of Föhn and temperature inversion) given the measurements taken from the Swiss meteorological monitoring network. The range of the methods used in the study includes data-driven feature selection, support vector algorithms and artificial neural networks.

Symmetric periodic orbits near heteroclinic loops at infinity for a class of polynomial vector fields

For polynomial vector fields in R3, in general, it is very difficult to detect the existence of an open set of periodic orbits in their phase portraits. Here, we characterize a class of polynomial vector fields of arbitrary even degree having an open set of periodic orbits. The main two tools for proving this result are, first, the existence in the phase portrait of a symmetry with respect to a plane and, second, the existence of two symmetric heteroclinic loops.

A Robust Multiple Feature Approach To Endpoint Detection In Car Environment Based On Advanced Classifiers

In this paper we propose an endpoint detection system based on the use of several features extracted from each speech frame, followed by a robust classifier (i.e Adaboost and Bagging of decision trees, and a multilayer perceptron) and a finite state automata (FSA). We present results for four different classifiers. The FSA module consisted of a 4-state decision logic that filtered false alarms and false positives. We compare the use of four different classifiers in this task. The look ahead of the method that we propose was of 7 frames, which are the number of frames that maximized the accuracy of the system. The system was tested with real signals recorded inside a car, with signal to noise ratio that ranged from 6 dB to 30dB. Finally we present experimental results demonstrating that the system yields robust endpoint detection.

Investigation of ICA algorithms for feature extraction of EEG signals in discrimination of Alzheimer disease

In this paper we present a quantitative comparisons of different independent component analysis (ICA) algorithms in order to investigate their potential use in preprocessing (such as noise reduction and feature extraction) the electroencephalogram (EEG) data for early detection of Alzhemier disease (AD) or discrimination between AD (or mild cognitive impairment, MCI) and age-match control subjects.

Is tumour size an underestimated feature in the current TNM system for malignancies of the pancreatic head?

BACKGROUND: As the long-term survival of pancreatic head malignancies remains dismal, efforts have been made for a better patient selection and a tailored treatment. Tumour size could also be used for patient stratification. METHODS: One hundred and fourteen patients underwent a pancreaticoduodenectomy for pancreatic adenocarcinoma, peri-ampullary and biliary cancer stratified according to: ≤20 mm, 21-34 mm, 35-45 mm and >45 mm tumour size. RESULTS: Patients with tumour sizes of ≤20 mm had a N1 rate of 41% and a R1/2 rate of 7%. The median survival was 3.4 years. N1 and R1/2 rates increased to 84% and 31% for tumour sizes of 21-34 mm (P = 0.0002 for N, P = 0.02 for R). The median survival decreased to 1.6 years (P = 0.0003). A further increase in tumour size of 35-45 mm revealed a further increase of N1 and R1/2 rates of 93% (P < 0.0001) and 33%, respectively. The median survival was 1.2 years (P = 0.004). Tumour sizes >45 mm were related to a further decreased median survival of 1.1 years (P = 0.2), whereas N1 and R1/2 rates were 87% and 20%, respectively. DISCUSSION: Tumour size is an important feature of pancreatic head malignancies. A tumour diameter of 20 mm seems to be the cut-off above which an increased rate of incomplete resections and metastatic lymph nodes must be encountered and the median survival is reduced.

Modelling ecological niches with support vector machines

1. The ecological niche is a fundamental biological concept. Modelling species' niches is central to numerous ecological applications, including predicting species invasions, identifying reservoirs for disease, nature reserve design and forecasting the effects of anthropogenic and natural climate change on species' ranges. 2. A computational analogue of Hutchinson's ecological niche concept (the multidimensional hyperspace of species' environmental requirements) is the support of the distribution of environments in which the species persist. Recently developed machine-learning algorithms can estimate the support of such high-dimensional distributions. We show how support vector machines can be used to map ecological niches using only observations of species presence to train distribution models for 106 species of woody plants and trees in a montane environment using up to nine environmental covariates. 3. We compared the accuracy of three methods that differ in their approaches to reducing model complexity. We tested models with independent observations of both species presence and species absence. We found that the simplest procedure, which uses all available variables and no pre-processing to reduce correlation, was best overall. Ecological niche models based on support vector machines are theoretically superior to models that rely on simulating pseudo-absence data and are comparable in empirical tests. 4. Synthesis and applications. Accurate species distribution models are crucial for effective environmental planning, management and conservation, and for unravelling the role of the environment in human health and welfare. Models based on distribution estimation rather than classification overcome theoretical and practical obstacles that pervade species distribution modelling. In particular, ecological niche models based on machine-learning algorithms for estimating the support of a statistical distribution provide a promising new approach to identifying species' potential distributions and to project changes in these distributions as a result of climate change, land use and landscape alteration.