Surrogate modeling approximation using a mixture of experts based on EM joint estimation

An automatic method to combine several local surrogate models is presented. This method is intended to build accurate and smooth approximation of discontinuous functions that are to be used in structural optimization problems. It strongly relies on the Expectation−Maximization (EM) algorithm for Gaussian mixture models (GMM). To the end of regression, the inputs are clustered together with their output values by means of parameter estimation of the joint distribution. A local expert is then built (linear, quadratic, artificial neural network, moving least squares) on each cluster. Lastly, the local experts are combined using the Gaussian mixture model parameters found by the EM algorithm to obtain a global model. This method is tested over both mathematical test cases and an engineering optimization problem from aeronautics and is found to improve the accuracy of the approximation.     

A generic framework for median graph computation based on a recursive embedding approach

The median graph has been shown to be a good choice to obtain a representative of a set of graphs. However, its computation is a complex problem. Recently, graph embedding into vector spaces has been proposed to obtain approximations of the median graph. The problem with such an approach is how to go from a point in the vector space back to a graph in the graph space. The main contribution of this paper is the generalization of this previous method, proposing a generic recursive procedure that permits to recover the graph corresponding to a point in the vector space, introducing only the amount of approximation inherent to the use of graph matching algorithms. In order to evaluate the proposed method, we compare it with the set median and with the other state-of-the-art embedding-based methods for the median graph computation. The experiments are carried out using four different databases (one semi-artificial and three containing real-world data). Results show that with the proposed approach we can obtain better medians, in terms of the sum of distances to the training graphs, than with the previous existing methods.     

Boosting support vector machines for imbalanced data sets

Real world data mining applications must address the issue of learning from imbalanced data sets. The problem occurs when the number of instances in one class greatly outnumbers the number of instances in the other class. Such data sets often cause a default classifier to be built due to skewed vector spaces or lack of information. Common approaches for dealing with the class imbalance problem involve modifying the data distribution or modifying the classifier. In this work, we choose to use a combination of both approaches. We use support vector machines with soft margins as the base classifier to solve the skewed vector spaces problem. We then counter the excessive bias introduced by this approach with a boosting algorithm. We found that this ensemble of SVMs makes an impressive improvement in prediction performance, not only for the majority class, but also for the minority class.     

Methodology for capturing and formalizing DFM Knowledge

Design for manufacturing (DFM) practices lead to more competitive products from the point of view of cost, development time and quality. However, the success of considering manufacturing issues during design process would be higher if manufacturing information was more readily available and designers needed less experience to select information relevant to DFM.     

Characterization of renal tumours based on Raman spectra classification

In this study, we propose to evaluate the potential of Raman spectroscopy (RS) to assess renal tumours at surgery. Different classes of Raman renal spectra acquired during a clinical protocol are discriminated using support vector machines classifiers. The influence on the classification scores of various preprocessing steps generally involved in RS are also investigated and evaluated in the particular context of renal tumour characterization. Encouraging results show the interest of RS to evaluate kidney cancer and suggest the potential of this technique as a surgical assistance during partial nephrectomy.     

Gambling versus Designing: Organizing for the Design of the Probability Space in the Energy Sector

The objective of this paper is to elucidate an organizational process for the design of generic technologies (GTs). While recognizing the success of GTs, the literature on innovation management generally describes their design according to evolutionary strategies featuring multiple and uncertain trials, resulting in the discovery of common features among multiple applications. This random walk depends on multiple market and technological uncertainties that are considered exogenous: as smart as he can be, the ‘gambler’ must play in a given probability space. However, what happens when the innovator is not a gambler but a designer, i.e., when the actor is able to establish new links between previously independent emerging markets and technologies? Formally speaking, the actor designs a new probability space. Building on a case study of two technological development programmes at the French Center for Atomic Energy, we present cases of GTs that correspond to this logic of designing the probability space, i.e. the logic of intentionally designing common features that bridge the gap between a priori heterogeneous applications and technologies. This study provides another example showing that the usual trial‐and‐learning strategy is not the only strategy to design GTs and that these technologies can be designed by intentionally building new interdependences between markets and technologies. Our main result is that building these interdependences requires organizational patterns that correspond to a ‘design of exploration’ phase in which multiple technology suppliers and application providers are involved in designing both the probability space itself and the instruments to explore and benefit from this new space.     

Building language resources for a Multi-Engine English-Filipino machine translation system

In this paper, we present the building of various language resources for a multi-engine bi-directional English-Filipino Machine Translation (MT) system. Since linguistics information on Philippine languages are available, but as of yet, the focus has been on theoretical linguistics and little is done on the computational aspects of these languages, attempts are reported here on the manual construction of these language resources such as the grammar, lexicon, morphological information, and the corpora which were literally built from almost non-existent digital forms. Due to the inherent difficulties of manual construction, we also discuss our experiments on various technologies for automatic extraction of these resources to handle the intricacies of the Filipino language, designed with the intention of using them for the MT system. To implement the different MT engines and to ensure the improvement of translation quality, other language tools (such as the morphological analyzer and generator, and the part of speech tagger) were developed.     

NodeTrix: a hybrid visualization of social networks

The need to visualize large social networks is growing as hardware capabilities make analyzing large networks feasible and many new data sets become available. Unfortunately, the visualizations in existing systems do not satisfactorily resolve the basic dilemma of being readable both for the global structure of the network and also for detailed analysis of local communities. To address this problem, we present NodeTrix, a hybrid representation for networks that combines the advantages of two traditional representations: node-link diagrams are used to show the global structure of a network, while arbitrary portions of the network can be shown as adjacency matrices to better support the analysis of communities. A key contribution is a set of interaction techniques. These allow analysts to create a NodeTrix visualization by dragging selections to and from node-link and matrix forms, and to flexibly manipulate the NodeTrix representation to explore the dataset and create meaningful summary visualizations of their findings. Finally, we present a case study applying NodeTrix to the analysis of the InfoVis 2004 coauthorship dataset to illustrate the capabilities of NodeTrix as both an exploration tool and an effective means of communicating results.