Carlos Francisco Moreno-García

Effectiveness of social marketing strategies to reduce youth obesity in European school-based interventions: a systematic review and meta-analysis

Context: The use of social marketing to modify lifestyle choices could be helpful in reducing youth obesity. Some or all of the 8 domains of the National Social Marketing Centre’s social marketing benchmark criteria (SMBC) are often used but not always defined in intervention studies. Objective: The aim of this review is to assess the effectiveness of European school-based interventions to prevent obesity relative to the inclusion of SMBC domains in the intervention. Data Sources: The PubMed, Cochrane, and ERIC databases were used. Study Selection: Nonrandomized and randomized controlled trials conducted from 1990 to April 2014 in participants aged 5 to 17 years were included. Data Extraction: After the study selection, the 8 domains of the SMBC were assessed in each included study. Results: Thirty-eight publications were included in the systematic review. For the meta-analysis, randomized controlled trials (RCTs) reporting body mass index or prevalence of overweight and obesity were considered. Eighteen RCTs with a total of 8681 participants included at least 5 SMBC. The meta-analysis showed a small standardized mean difference in body mass index of −0.25 (95%CI, −0.45 to −0.04) and a prevalence of overweight and obesity odds ratio of 0.72 (95%CI, 0.5–0.97). Conclusion: Current evidence indicates that the inclusion of at least 5 SMBC domains in school-based interventions could benefit efforts to prevent obesity in young people. PROSPERO registration number: CRD42014007297.

A Graph Repository for Learning Error-Tolerant Graph Matching

In the last years, efforts in the pattern recognition field have been especially focused on developing systems that use graph based representations. To that aim, some graph repositories have been presented to test graph-matching algorithms or to learn some parameters needed on such algorithms. The aim of these tests has always been to increase the recognition ratio in a classification framework. Nevertheless, some graph-matching applications are not solely intended for classification purposes, but to detect similarities between the local parts of the objects that they represent. Thus, current state of the art repositories provide insufficient information. We present a graph repository structure such that each register is not only composed of a graph and its class, but also of a pair of graphs and a ground-truth correspondence between them, as well as their class. This repository structure is useful to analyse and develop graph-matching algorithms and to learn their parameters in a broadly manner. We present seven different databases, which are publicly available, with these structure and present some quality measures experimented on them.

Consensus of multiple correspondences between sets of elements

A method to deduct a consensus correspondence given some correspondences.We explain in detail 3 methods: Voting, Iterative and Agglomerative.A complete evaluation on known datasets. In many pattern recognition and computer vision problems, it is often necessary to compare multiple sets of elements that are completely or partially overlapping and possibly corrupted by noise. Finding a correspondence between elements from the different sets is one of the crucial tasks that several computer vision, robotics or image registration methods have to cope with. The aim of this paper is to find a consensus correspondence between two sets of points, given several initial correspondences between these two sets. We present three different methods: iterative, voting and agglomerative. If the noise randomly affects the original data, we suppose that, while using the deducted correspondence, the process obtains better results than each individual correspondence. The different correspondences between two sets of points are obtained through different feature extractors or matching algorithms. Experimental validation shows the runtime and accuracy for the three methodologies. The agglomerative method obtains the highest accuracy compared to the other consensus methods and also the individual ones, while obtaining an acceptable runtime.

Correspondence consensus of two sets of correspondences through optimisation functions

AbstractWe present a consensus method which, given the two correspondences between sets of elements generated by separate entities, enounces a final correspondence consensus considering the existence of outliers . Our method is based on an optimisation technique that minimises the cost of the correspondence while forcing (to the most) to be the mean correspondence of the two original correspondences. The method decides the mapping of the elements that the original correspondences disagree and returns the same element mapping when both correspondences agree. We first show the validity of the method through an experiment in ideal conditions based on palmprint identification, and subsequently present two practical experiments based on image retrieval.

Partial to Full Image Registration Based on Candidate Positions and Multiple Correspondences

In some image-registration based applications, it is more usual to detect a low quality and tiny partial image rather than a full sample (forensic palmprint recognition, satellite images, object detection in outdoor scenes …). In these cases, the usual registration methods fail due to the great amount of outliers that have to be detected while comparing a tiny image (object to be registered) to a full image (object in the database). In this paper, we present an image registration method that explicitly considers a great amount of outliers. In a first step, the method selects some candidate points to be the centres of the partial image. In a second step, these candidates are refined until selecting one through a multiple correspondence method. Experimental validation shows that the algorithm can outperform state of the art identification methods given the image to be identified a tiny and partial sample.

Weighted Mean Assignment of a Pair of Correspondences Using Optimisation Functions

Consensus strategies have been recently studied to help machine learning ensure better results. Likewise, optimisation in graph matching has been explored to accelerate and improve pattern recognition systems. In this paper, we present a fast and simple consensus method which, given two correspondences of sets generated by separate entities, enounces a final consensus correspondence. It is based on an optimisation method that minimises the cost of the correspondence while forcing it to the most to be a weighted mean. We tested our strategy comparing ourselves with the classical minimum cost matching system, using a palmprint database, with each palmprint is represented by an average of 1000 minutiae.

Iterative Versus Voting Method to Reach Consensus Given Multiple Correspondences of Two Sets

Point Set Registration is the process of finding the correspondence between points of two sets. There are some Point Set Registration applications in which given two sets of points, several correspondences between these points may be deducted. However, the use of different parameters or optimisation strategies makes these correspondences differ from each other. In this paper, we present two different methods to obtain a consensus correspondence given several correspondences between two sets of points. The first one is based on the classical voting strategy. The second one iteratively updates the consensus correspondence given two correspondences: a non-previously explored corre- spondence and the current consensus. In this last method, the computation of the consensus given two correspondences is done through a method recently pub- lished. We compare the voting and iterative methods using an image dataset and validate the runtime and the quality of the consensus correspondence using an existing homography between the considered images.

Generalised Median of a Set of Correspondences Based on the Hamming Distance

A correspondence is a set of mappings that establishes a relation between the elements of two data structures (i.e. sets of points, strings, trees or graphs). If we consider several correspondences between the same two structures, one option to define a representative of them is through the generalised median correspondence. In general, the computation of the generalised median is an NP-complete task. In this paper, we present two methods to calculate the generalised median correspondence of multiple correspondences. The first one obtains the optimal solution in cubic time, but it is restricted to the Hamming distance. The second one obtains a sub-optimal solution through an iterative approach, but does not have any restrictions with respect to the used distance. We compare both proposals in terms of the distance to the true generalised median and runtime.

On the Influence of Node Centralities on Graph Edit Distance for Graph Classification

Classical graph approaches for pattern recognition applications rely on computing distances between graphs in the graph domain. That is, the distance between two graphs is obtained by directly optimizing some objective function which consider node and edge attributes. Bipartite Graph Matching was first published in a journal in 2009 and new versions have appeared to speed up its runtime such as the Fast Bipartite Graph Matching. This algorithm is based on defining a cost matrix between all nodes of both graphs and solving the node correspondence through a linear assignment method. To construct the matrix, several local structures can be defined from the simplest one (only the node) to the most complex (a whole clique or eigenvector structure). In this paper, we propose five different options and we show that the type of local structure and the distance defined between these structures is relevant for graph classification.

An Edit Distance Between Graph Correspondences

The Hamming Distance has been largely used to calculate the dissimilarity of a pair of correspondences (also known as labellings or matchings) between two structures (i.e. sets of points, strings or graphs). Although it has the advantage of being simple in computation, it does not consider the structures that the correspondences relate. In this paper, we propose a new distance between a pair of graph correspondences based on the concept of the edit distance, called Correspondence Edit Distance. This distance takes into consideration not only the mapped elements of the correspondences, but also the attributes on the nodes and edges of the graphs being mapped. In addition to its definition, we also present an efficient procedure for computing the correspondence edit distance in a special case. In the experimental validation, the results delivered using the Correspondence Edit Distance are contrasted against the ones of the Hamming Distance in a case of finding the weighted means between a pair of graph correspondences.