Search

Breast cancer disease is considered to be the second leading reason for death among women. Unfortunately, even if the treatment of cancer started soon after diagnosis, the cancer cells may remain in the body, and cancer may recur. Various Machine Learning (ML) methods to predict breast cancer recurrence were applied recently, and the ML methods’ performance needs to be examined to determine the proper method for prediction. Usually, the datasets contain many features which may sometimes mislead the prediction process; as some features may lead to confusion or inaccurate prediction. Thus, in this study, two breast cancer recurrence datasets were statistically analyzed and further refined by Brain Storming Optimization algorithm (BSO). The proposed multi-stages technique consists of three main stages; first, the statistical feature selection methods (SFM) which statistically select the discriminative features based on importance ranking and features correlations and statistical hypothesis testing, to be passed to the second stage. The features are ranked based on their correlation with the class variable. The second stage namely the multi classifier (MC), which evaluates each method based on three classifiers and produces a combination of features performed by two SFM and three classifiers. In the third stage, the best combination of the selected features was recognized by the BSO algorithm to search for an optimal solution that produces the highest accuracy. In addition, the BSO algorithm has been modified to deal with the feature selection problem. The performance of the proposed technique was evaluated by stratified 10-fold cross-validation. As a result, the multi-stage learning technique showed to be effective in ranking the features and improved the classification accuracy for breast cancer recurrence.

The Oil and Gas (O&G) industry is used to cycles of lows and highs due to different challenging economic and political situations. Yet the challenges caused by the sanitary crisis due to the covid-19 pandemic are certainly like no others. The shutdown of a large number of social activities had a direct impact on energy consumption. Many studies [1], [2] and [3] have been published at the beginning of the covid-19 pandemic to predict impacts of the restrictions imposed on a global scale: decline in global oil demand, saturation of storage capacities and delay of exploration and production projects. Companies facing this unprecedented crisis had no option but to adopt innovative ways of driving costs lower and maximizing operational efficiency. As a consequence, the pace at which Data Science (DS) is finding its way to O&G applications has been noticeably accelerated although the O&G industry is one of the latecomers to digitalization [4]. The adoption of DS and data-driven solutions has moved from gaining acceptance in the industry to becoming a necessity to many companies. According to a Gartner survey [5], the O&G industry commitment to investment in digital transformation in general had become the first priority in 2021 while it was third-highest priority in 2019 and not even funded in 2014. This involves investments in data acquisition techniques through innovative sensing technologies but also investments in advanced data aggregation and analytics platforms. AI/ML/analytics are listed in the same survey [5] as "top game-changing technologies in 2021". The 2021 survey also states that 50% of the O&G companies have plans to increase their investments in AI/ML and related fields such as cloud-computing. But adoption and operationalization of DS does not come with no challenges. Acceptance and reliance on data-driven models need a favorable cultural and technical environment that is not necessarily compatible with the conventional corporate-like outlook of O&G companies: Data privacy and ownership regulations can diminish DS efforts. Security restrictions can prevent deployment of ML models to end users. All of these challenges are accentuated by the absence of a clear process model to implement and manage DS projects. In this paper, we survey the actual challenges the O&G industry is facing and present a number of corresponding solutions. The paper is structured as follows. The first section explores the state of the art of data-driven models in the O&G industry. The second section lists the challenges DS is facing within the O&G industry and proposes a classification of these challenges into three main classes, namely: human, data and infrastructure related challenges. The paper also proposes an O&G specific framework for DS projects to overcome these identified challenges.

Chronic degradation of hybrid layer collagen by matrix metalloproteinases (MMPs) jeopardizes resin-dentin interfacial integrity and limits the durability of dental restorations. The 4-point bending strength (BS) is a valid but uncommon method of testing the mechanical behavior of resin-dentin interfaces. The present study aims to analyze the influence of two matrix metalloproteinase inhibitors on microtensile bond strength (µTBS), BS and nanoleakage. A total of 48M were divided into three groups according to bonding procedure. Teeth were horizontally sectioned to produce a flat dentin surface. In the control group, etch-and-rinse PrimeBond One (Dentsply) bonding was used; in the self-etch group, methacryloyloxydodecylpyridinium bromide (MDPB)-containing Clearfil SE Protect (Kuraray) was used; and in the benzalkonium chloride (BAC)-etch group, BAC-etchant (Bisco) was used. A Ceram.X-One (Dentsply) composite was built as three successive layers and was light-cured. Samples were sectioned to produce microrods that were randomly divided into two groups for analysis at baseline and after 6 months of water immersion (n = 32), plus one slab for nanoleakage analysis (n = 8) via scanning electron microscopy (SEM) and digital image analysis (Fiji). Data were analyzed using the Weibull distribution and a mixed-model ANOVA with a post hoc Tukey test. All groups showed deterioration of the initial bonds. The self-etch group had a worse baseline µTBS than the control but had the best BS after aging. BAC-etch did not improve bond stability of etch-and-rinse adhesive. The µTBS and BS test results after aging were moderately correlated. Mixed fractures prevailed with regard to µTBS, whereas adhesive fractures dominated with regard to BS. Nanoleakage was not eliminated in any group and increased after aging. MDPB self-etch resisted bond degradation better than etch-and-rinse adhesives, even after BAC-etching. Integrating BS in studies of µTBS and nanoleakage might provide more clinically relevant outcomes for predicting the performance of dental adhesives.

We present a framework for road network management to assist road authorities in maintenance budget estimations and long-term maintenance strategies definition. Information about road conditions is obtained from monitoring. Available data are used to estimate and update prediction of degradation evolution. Our decision-making under uncertainty tool is based on a Markov decision process to minimize total expected costs not only from a financial perspective but also from social and environmental ones. This is achieved by combining both technical/condition indicators and performance indicators, i.e., end-user service levels, in the problem formulation. From a practical perspective, our tool can be adapted to different road authority's needs. In this paper we highlight three applications. First, the tool can be used to select and develop new, cost-effective monitoring techniques that sensor relevant condition indicators or end-user service levels. Second, the tool provides decision rules for efficient maintenance strategies as well as long-term policies for budget estimation. Finally, increasing budget restrictions, funding cuts are expected to increase future budgets, but also to have negative social and environmental impacts. This tool can be used to assess impacts of budget cuts. © 2014 Taylor & Francis Group, London..

International audience; Classical results in maintenance optimization are based on the characterization of an average system degradation behaviour. The main reasons of such approaches are the robustness of the decisions with respect to the system state and the time horizon. By cons, the ''smoothing'' effect in existing degradation models and the ''generic'' aspect of the decision affect the decision quality as these models do not take into account potential changes in deterioration modes. Such consideration can be particularly valuable in a safety context. We propose here a condition-based maintenance approach which adapts the maintenance decision according to the degradation behaviour updated using current state observations. First, we propose to extend the system state definition, usually limited to an indicator of observable degradation by adding information related to the speed of deterioration referred hereafter as potential of degradation or deterioration growth rate. Moreover, we propose to take benefit of current observations as well as performed actions to update the degradation laws. More than the model fitness improvement, another advantage of our approach is the proposition of a more realistic modelling of the maintenance impact onto the future behaviour of the system. However, the introduction of the potential of degradation as a new decision parameter does not prejudge an intuitive structure for the decision policy. Moreover, as the classical concept of the failure rate, the potential of degradation is non observable. which implies more efforts in both optimization modelling and solution procedure. We highlight the structural properties of the optimization problem which ensure an optimal control limit policy. We will conclude our communication by the illustration of the results of our maintenance model in a pavement management context.

International audience; A condition-based maintenance optimization approach is developed for the road-cracking problem in order to derive optimal maintenance policies that minimize a total discounted maintenance cost. The approach is based on a Markov decision process that takes into ac- count multiple actions with varying effects on future road performance. Maintaining the road consists of adding a new asphalt layer; however, as resurfacing actions are constrained by a maximum total road thickness, the maintenance decision is not only how thick a layer to apply, but also how much old road to remove. Each combination of these actions leads to different maintenance costs and different future degradation behaviours. The road state is modelled by a dependent bivariate deterioration variable (the longitudinal cracking percentage and the deterioration growth rate), for taking these different changes in the cracking patterns into account. Moreover, the sensitivity to cracking for existing roads can be reduced with the addition of new layers, and thus actions that can lead to states better than good-as-new have to be considered. A numerical analysis is provided to illustrate the benefits of the introduction of the deterioration speed in the decision framework, as well as the belief that initially building a road to its maximum thickness is not optimal. The trade-offs in the design decisions and the exploitation/maintenance costs are also explored.

International audience; We develop an optimal maintenance policy for a road section to minimize the total maintenance cost over the infinite horizon when some deterioration and decision parameters are not observable. Both perfect and imperfect maintenance actions are possible through the application of various thicknesses of resurfacing layers. We use a two-phase deterioration process based on two parameters: the longitudinal cracking percentage and the deterioration growth rate. Our deterioration model is a state-based model based on the state-dependent Gamma process for the longitudinal cracking percentage and the Bilateral Gamma process for the deterioration growth rate. Moreover the maintenance decision is constrained by a maximum road thickness that makes the maintenance decisions more complex as it becomes how much surface layer to add as well as to remove. Because only one of the two deterioration parameters is observable, we formulate the problem as a partially observed Markov decision process and solve it using a grid-based value iteration algorithm. Numerical examples have shown that our model provides a preventive maintenance policy that slows down the initiation as well as the propagation of longitudinal cracks and that may ameliorate the road state to a better than as-good-as-new one by altering its composition through additive resurfacing layers.

International audience; We propose a maintenance decision framework where the state of the system is based on a percentage of total degrada- tion and the rate of change of the degradation. At each decision interval, the degradation percentage may be observed perfectly as well as the rate of change with respect to the last observation. We consider five possible actions based on these two observable phenomenon: Do nothing, three imperfect maintenance actions, and a complete renewal. Prior imperfect maintenance models do not consider a scenario where imperfect maintenance not only restores the system to a state less than “as good as new” but also under a new deterioration law. In this case, it is necessary to consider that the rate of degradation after imperfect maintenance will be altered based on the action performed and the state of the system prior to the maintenance. We propose to develop an imperfect maintenance optimization model based on a Markov decision process framework to determine the optimal maintenance actions given the state of the system at each decision interval and motivate the problem through road maintenance.

We develop a condition-based maintenance optimization approach for the road cracking problem in order to derive an optimal action-planning policy from a Markov decision process that minimizes the expected cost. Our model deals with multiple imperfect actions that consist of different resurfacing thicknesses and that have varying effects on the future deterioration law. We model the deterioration after maintenance to be dependent upon both the state of the road before maintenance and the type of maintenance performed. Another new aspect of our model is the possibility for a maintenance action to render the road to a state better than as-good-as-new. The application of new road layers is, however, constrained by a maximum total road thickness. Some maintenance actions may then become infeasible due to the thickness of the road relative to the constraint. As the road is constrained by a maximum total thickness, the maintenance decision is made complex by the determination of not only how thick of a layer to apply, but also how much old road to remove. We model the road state by two deterioration variables: the longitudinal cracking percentage and its associated growth rate. Our degradation model will take into account the changing road composition which is a function of the successive application of new and removal of old layers.

This work is a part of a research project for the development of new condition-based approaches for road maintenance optimization. The degradation pattern on interest is here the longitudinal cracking process due to cumulative fatigue (the traffic repetition leads to the occurrence of cracks in the road basement which grows up to the road surface). In this context, we have proposed a new theoretical deterioration model based on two dependent indicators -the observable deterioration measurement and the potential deterioration growth. Even if the construction of the model is based on practical considerations, its application in an operational context remains difficult. The objective of this communication is to study the applicability of such a model and propose some improvements. After analyzing the original model to highlight its strengths and limitations, we propose to revisit the definitions of the decision parameters while specifying the construction of the associated functions. A statistical inference procedure is discussed. A numerical example based on the original maintenance model is presented to illustrate the benefits of this approach that we will present as some “best practices” for future road pavement management.