Luis A. A. Meira

Selection of Representative Models for Decision Analysis Under Uncertainty

The decision-making process in oil fields includes a step of risk analysis associated with the uncertainties present in the variables of the problem. Such uncertainties lead to hundreds, even thousands, of possible scenarios that are supposed to be analyzed so an effective production strategy can be selected. Given this high number of scenarios, a technique to reduce this set to a smaller, feasible subset of representative scenarios is imperative. The selected scenarios must be representative of the original set and also free of optimistic and pessimistic bias. This paper is devoted to propose an assisted methodology to identify representative models in oil fields. To do so, first a mathematical function was developed to model the representativeness of a subset of models with respect to the full set that characterizes the problem. Then, an optimization tool was implemented to identify the representative models of any problem, considering not only the cross-plots of the main output variables, but also the risk curves and the probability distribution of the attribute-levels of the problem. The proposed technique was applied to two benchmark cases and the results, evaluated by experts in the field, indicate that the obtained solutions are richer than those identified by previously adopted manual approaches. The program bytecode is available under request. HighlightsA new optimization-based method to select representative models in oil fields.A new mathematical function that captures the representativeness of a set of models.The mathematical function is combined with an optimization metaheuristic.The proposal was applied to the UNISIM-I-D benchmark problem to validate the methodology.Experts indicate that results are richer than those obtained by other approaches.

Acc-Motif: accelerated network motif detection

Network motif algorithms have been a topic of research mainly after the 2002-seminal paper from Milo et al. [1], which provided motifs as a way to uncover the basic building blocks of most networks. Motifs have been mainly applied in Bioinformatics, regarding gene regulation networks. Motif detection is based on induced subgraph counting. This paper proposes an algorithm to count subgraphs of size k + 2 based on the set of induced subgraphs of size k. The general technique was applied to detect 3, 4 and 5-sized motifs in directed graphs. Such algorithms have time complexity O(a(G)m), O(m2) and O(nm2), respectively, where a(G) is the arboricity of G(V, E). The computational experiments in public data sets show that the proposed technique was one order of magnitude faster than Kavosh and FANMOD. When compared to NetMODE, acc-Motif had a slightly improved performance.

HOW FAR do WE GET USING MACHINE LEARNING BLACK-BOXES?

With several good research groups actively working in machine learning (ML) approaches, we have now the concept of self-containing machine learning solutions that oftentimes work out-of-the-box leading to the concept of ML black-boxes. Although it is important to have such black-boxes helping researchers to deal with several problems nowadays, it comes with an inherent problem increasingly more evident: we have observed that researchers and students are progressively relying on ML black-boxes and, usually, achieving results without knowing the machinery of the classifiers. In this regard, this paper discusses the use of machine learning black-boxes and poses the question of how far we can get using these out-of-the-box solutions instead of going deeper into the machinery of the classifiers. The paper focuses on three aspects of classifiers: (1) the way they compare examples in the feature space; (2) the impact of using features with variable dimensionality; and (3) the impact of using binary classifiers to solve a multi-class problem. We show how knowledge about the classifiers machinery can improve the results way beyond out-of-the-box machine learning solutions.

A continuous facility location problem and its application to a clustering problem

We consider a new problem, which we denote by Continuous Facility Location (ConFL), and its application to the k-Means Problem. Problem ConFL is a natural extension of the Uncapacitated Facility Location Problem where a facility can be any point in Rq. The proposed algorithms are based on a primal-dual technique for spaces with constant dimensions. For the ConFL Problem, we present algorithms with approximation factors 3 + ε and 1.861 + ε for euclidean distances and 9 + ε for squared euclidean distances. For the k-Means Problem (that is restricted to squared euclidean distance), we present an algorithm with approximation factor 54+ ε. All algorithms have good practical behaviour in small dimensions. Comparisons with known algorithms show that the proposed algorithms have good practical behaviour.

Accelerated Motif Detection Using Combinatorial Techniques

Network motif algorithms have been a topic of research mainly after the 2002-seminal paper from Milo et al, that provided motifs as a way to uncover the basic building blocks of most networks. This article proposes new algorithms to exactly count isomorphic pattern motifs of size 3 and 4 in directed graphs. The algorithms are accelerated by combinatorial techniques. Let G(V, E) be a directed graph with m=|E|. We describe an O(m√m) time complexity algorithm to count isomorphic patterns of size 3. To counting isomorphic patterns of size 4, we propose an O(m2) algorithm. The new algorithms were implemented and compared with Fanmod motif detection tool. The experiments show that our algorithms are expressively faster than Fanmod. We also let our tool to detect motifs, the ACC-MOTIF, available in the Internet.

RMFinder 2.0: An Improved Interactive Multi-Criteria Scenario Reduction Methodology

This paper presents an extension of the RMFinder technique, previously proposed to identify representative models (RMs) within the decision-making process in oil fields. As there are several uncertainties associated with this decision-making process, a large number of scenarios are supposed to be analyzed, so that high-quality production strategies can be defined. Such broad analysis is often unfeasible, so techniques to automatically identify RMs are particularly relevant. The original RMFinder does not consider the individual probability of each RM, which may not be accurate when the risk curves of the problem are estimated. Therefore, a mechanism to calculate the individual probability of each RM was developed here, together with a graphical way to visualize different proposals of RMs. To automatically identify the optimal probability of each RM, this new version of RMFinder minimizes the deviation between the risk curves generated with the selected RMs and the original risk curves of the problem. The graphical approach automatically exhibits, in a single page per solution, the RM dispersion in the scatter plots, the resulting risk curves and the differences between attribute-level distributions. This helps the decision makers to visualize and compare different sets of RMs. The proposed methodology was applied to a small synthetic problem and to three reservoir models based on real-world Brazilian fields: (i) UNISIM-I-D, a benchmark case based on the Namorado field; (ii) UNISIM-II-D, a benchmark case based on a highly fractured pre-salt carbonate reservoir; and (iii) ST001a, a highly heterogeneous heavy oil offshore field. The obtained sets of RMs were evaluated by experts and considered appropriate to the studied problems, being adopted as the standard models in the following steps of the decision-making process to define the production strategies under uncertainties. Introduction One of the main goals of the decision making process in oil fields is a proper definition of the production strategy to be implemented. The basis of the production strategy definition procedure is the geological model of the field, which is used by computational simulators to estimate the behavior of the reservoir during the concession time (given a production strategy). Therefore, given the behavior of the reservoir, it is possible to evaluate the quality of the production strategy. Notice that the production strategies can be evaluated according to different criteria, such as the Net Present Value (NPV), for example. In this context, geological models of oil fields usually contain 1 The production strategy is composed by the infrastructure of wells, pipes, platforms and manifolds, for example. All manuscripts will be sent through an XML tagging process that will standardize the look of the paper and create links for figures, equations, and references. Figures and tables should be placed directly after the first paragraph they are mentioned in. The XML tagging will not alter the technical content of the paper.

How Far You Can Get Using Machine Learning Black-Boxes

Supervised Learning (SL) is a machine learning research area which aims at developing techniques able to take advantage from labeled training samples to make decisions over unseen examples. Recently, a lot of tools have been presented in order to perform machine learning in a more straightforward and transparent manner. However, one problem that is increasingly present in most of the SL problems being solved is that, sometimes, researchers do not completely understand what supervised learning is and, more often than not, publish results using machine learning black-boxes. In this paper, we shed light over the use of machine learning black-boxes and show researchers how far they can get using these out-of-the-box solutions instead of going deeper into the machinery of the classifiers. Here, we focus on one aspect of classifiers namely the way they compare examples in the feature space and show how a simple knowledge about the classifier’s machinery can lift the results way beyond out-of-the-box machine learning solutions.

A greedy approximation algorithm for the uniform metric labeling problem analyzed by a primal-dual technique

We consider the uniform metric labeling problem. This NP-hard problem considers how to assign objects to labels respecting assignment and separation costs. The known approximation algorithms are based on solutions of large linear programs and are impractical for moderate- and large-size instances. We present an 8log n-approximation algorithm that can be applied to large-size instances. The algorithm is greedy and is analyzed by a primal-dual technique. We implemented the presented algorithm and two known approximation algorithms and compared them at randomized instances. The gain of time was considerable with small error ratios. We also show that the analysis is tight, up to a constant factor.

A Greedy Approximation Algorithm for the Uniform Labeling Problem Analyzed by a Primal-Dual Technique

In this paper we present a new fast approximation algorithm for the Uniform Metric Labeling Problem. This is an important classification problem that occur in many applications which consider the assignment of objects into labels, in a way that is consistent with some observed data that includes the relationship between the objects.

Semidefinite Programming Based Algorithms for the Sparsest Cut Problem

In this paper we analyze a known relaxation for the Spars- est Cut problem based on positive semidefinite constraints, and we present a branch and bound algorithm and heuristics based on this relaxation. The relaxed formulation and the algorithms were tested on small and moderate sized instances. It leads to values very close to the optimum solution values. The exact algorithm could obtain solutions for small and moderate sized instances, and the best heuristics obtained optimum or near optimum solutions for all tested instances. The semidefinite relaxation gives a lower bound C/W and each heuristic produces a cut S with a ratio c S /w S , where either c S is at most a factor of C or w S is at least a factor of W. We solved the semidefinite relaxation using a semi-infinite cut generation with a commercial linear programming package adapted to the sparsest cut problem. We showed that the proposed strategy leads to a better performance compared to the use of a known semidefinite programming solver.