Search

Background Ab initio prediction of splice sites is an essential step in eukaryotic genome annotation. Recent predictors have exploited Deep Learning algorithms and reliable gene structures from model organisms. However, Deep Learning methods for non-model organisms are lacking. Results We developed Spliceator to predict splice sites in a wide range of species, including model and non-model organisms. Spliceator uses a convolutional neural network and is trained on carefully validated data from over 100 organisms. We show that Spliceator achieves consistently high accuracy (89–92%) compared to existing methods on independent benchmarks from human, fish, fly, worm, plant and protist organisms. Conclusions Spliceator is a new Deep Learning method trained on high-quality data, which can be used to predict splice sites in diverse organisms, ranging from human to protists, with consistently high accuracy.

In the field of Reinforcement Learning, models based on neural networks are highly performing, but explaining their decisions is very challenging. Instead of seeking to open these "black boxes" to meet the increasing demand for explainability, another approach is to used rule-based machine learning models that are explainable by design, such as the Anticipatory Learning Classifier Systems (ALCS). ALCS are able to develop simultaneously a complete representation of their environment and a decision policy based on this representation to solve their learning tasks. This paper focuses on the ability of ALCS to deal with non-deterministic environments used in reinforcement learning problems, while discussing their explainability. Directions for future research are thus highlighted to improve both the performance and the explainability of the ALCS to meet the needs of critical real-world applications.

In this paper we consider diffusions on the half line ( 0 , ∞ ) (0,\infty ) such that the expectation of the arrival time at the origin is uniformly bounded in the initial point. This implies that there is a well defined diffusion process starting from infinity which takes finite values at positive times. We study the behavior of hitting times of large barriers and, in a dual way, the behavior of the process starting at infinity for small time. In particular, we prove that the process coming down from infinity is in small time governed by a specific deterministic function. Suitably normalized fluctuations of the hitting times are asymptotically Gaussian. We also derive the tail of the distribution of the hitting time of the origin and a Yaglom limit for the diffusion starting from infinity. We finally prove that the distribution of this process killed at the origin is absolutely continuous with respect to the speed measure. The density is expressed in terms of the eigenvalues and eigenfunctions of the generator of the killed diffusion.

International audience; Bioenergy systems are promoted in an effort to mitigate climate change, and policies are defined accordingly to be implemented in the coming decades. Life Cycle Assessment (LCA) is used to assess the environmental performance of bioenergy systems, yet subject to the limitations of static approaches. In classical LCA, no temporal differentiation is undertaken: all inventoried instant to long-term greenhouse gases emissions (GHG) are aggregated and characterised in the same way, over a fixed time horizon, by means of fixed characterisation factors. Positive and negative impact contributions of dynamic biogenic carbon (Cbio) sum up to zero, yielding the same result as carbon neutral estimates. Climate mitigation results are biased without the temporal consideration of these flows. The purpose of the study is to highlight the time-sensitive potential climatic consequences of policy-driven transport strategies for metropolitan France, in the specific context of the dynamic LCA framework and climate change mitigation. We therefore propose a dynamic approach coupling a partial-equilibrium model (PEM) with dynamic Cbio models. The PEM analyses in detail the techno-economic performance of the metropolitan French energy-transport sector. It explores prospective optimization options (supply-demand equilibrium) of emerging commodity and energy process pathways in response to a policy in question. The Cbio model generates dynamic inventories of the Cbio embedded in the primary renewable biomass outputs of the PEM. It captures the dynamic Cbio exchange flows between the atmosphere and the technosphere over time: negative emissions from fixation (sequestration) and positive emissions from release (e.g. combustion or decay). A dynamic impact method is applied to evaluate the mitigation effects of Cbio from forest wood residues by comparing the climate change impacts from complete carbon (fossil + biogenic) with carbon neutral inventories across scenarios. Two sets of results are computed concerning the overall transport (all emissions) and bioethanol (wood-to-fuel emissions) systems. The mitigation effect from long-term historic sequestration allocated to bioethanol (462%) is significantly larger than for transport (3%), expressed as the difference with carbon neutral estimates. The fossil-sourced emissions from bioethanol production represents only 5.4%. In contrast, a comparison with an alternative reference scenario involving wood decay demonstrated higher impacts (i.e. an increase of 316%) than carbon neutral estimates. The representation of the actual climatic consequences depends on the chosen fixed end-year of the dynamic impact assessment. Moreover, the mitigation effect is proven sensitive to the rotation length of forestry wood: the shorter the length the lower the mitigation from using renewable forest resources. Other energy-policy scenarios, Cbio modelling approaches and consequences of indirect effects should be further studied and contrasted.

I will shortly review the history of experimental and theoretical findings on period doubling until the discovery of the quantitative universal properties of the infinite period-doubling cascade.

Magnetic refrigeration (MR) is an alternative technology to conventional vapour compression with a high potential to reduce energy consumption and greenhouse gases. The working principle of MR is based on the property of magneto caloric materials (MCMs) to respond to applied external magnetic field by a variation in their temperature. To find the inexpensive MCMs with needed physical properties is still an issue. This paper describes a new method, whose objective is to automate the characterization process of MCM physical properties. This method is based on a multi-objective evolutionary optimization to define the combination of Hamiltonian model free parameters, which corresponds to needed physical properties of MCM. The advantage of this method is that it doesn’t need first-principles calculations or huge databases. The description is followed by a thorough validation of the proposed method for two alloys. The obtained results are in good agreement with the experimental measurements.

We consider spin-flip dynamics of configurations in $\{-1,1\}^{\mathbb{Z}^d}$, and study the time evolution of concentration inequalities. For "weakly interacting" dynamics we show that the Gaussian concentration bound is conserved in the course of time and it is satisfied by the unique stationary Gibbs measure. Next we show that, for a general class of translation-invariant spin-flip dynamics, it is impossible to evolve in finite time from a low-temperature Gibbs state towards a measure satisfying the Gaussian concentration bound. Finally, we consider the time evolution of the weaker uniform variance bound, and show that this bound is conserved under a general class of spin-flip dynamics., Comment: 25 pages, to appear in J. Stat. Phys. (2021)

International audience; PurposeConsequential life cycle assessment (C-LCA) aims to assess the environmental consequences of a decision. It differs from traditional LCA because its inventory includes all the processes affected by the decision which are identified by accounting for causal links (physical, economic, etc.). However, C-LCA results could be quite uncertain which makes the interpretation phase harder. Therefore, strategies to assess and reduce uncertainty in C-LCA are needed. Part of uncertainty in C-LCA is due to spatial variability that can be reduced using regionalization. However, regionalization can be complex and time-consuming if straightforwardly applied to an entire LCA model.MethodsThe main purpose of this article is to prioritize regionalization efforts to enhance interpretation in C-LCA by assessing the spatial uncertainty of a case study building on a partial equilibrium economic model. Three specific objectives are derived: (1) perform a C-LCA case study of alternative transportation scenarios to investigate the benefits of implementing a public policy for energy transition in France by 2050 with an uncertainty analysis to explore the strength of our conclusions, (2) perform global sensitivity analyses to identify and quantify the main sources of spatial uncertainty between foreground inventory model from partial equilibrium economic modeling, background inventory model and characterization factors, (3) propose a strategy to reduce the spatial uncertainty for our C-LCA case study by prioritizing regionalization.Results and discussionResults show that the implementation of alternative transport scenarios in compliance with public policy for the energy transition in France is beneficial for some impact categories (ICs) (global warming, marine acidification, marine eutrophication, terrestrial acidification, thermally polluted water, photochemical oxidant formation, and particulate matter formation), with a confidence level of 95%. For other ICs, uncertainty reduction is required to determine conclusions with a similar level of confidence. Input variables with spatial variability from the partial equilibrium economic model are significant contributors to the C-LCA spatial uncertainty and should be prioritized for spatial uncertainty reduction. In addition, characterization factors are significant contributors to the spatial uncertainty results for all regionalized ICs (except land occupation IC).ConclusionsWays to reduce the spatial uncertainty from economic modeling should be explored. Uncertainty reduction to enhance the interpretation phase and the decision-making should be prioritized depending on the goal and scope of the LCA study. In addition, using regionalized CFs in C-LCA seems to be relevant, and C-LCA calculation tools should be adapted accordingly.; ObjectifL'analyse du cycle de vie conséquentielle (ACV-C) vise à évaluer les conséquences environnementales d'une décision. Elle diffère de l'ACV traditionnelle car son inventaire comprend tous les processus affectés par la décision qui sont identifiés en tenant compte des liens de causalité (physiques, économiques, etc.). Cependant, les résultats de l'ACV-C peuvent être assez incertains, ce qui rend la phase d'interprétation plus difficile. Par conséquent, des stratégies pour évaluer et réduire l'incertitude dans l'ACV-C sont nécessaires. Une partie de l'incertitude de l'ACC est due à la variabilité spatiale qui peut être réduite en utilisant la régionalisation. Cependant, la régionalisation peut être complexe et prendre beaucoup de temps si elle est appliquée directement à un modèle d'ACV complet.MéthodesL'objectif principal de cet article est de prioriser les efforts de régionalisation pour améliorer l'interprétation dans l'ACV-C en évaluant l'incertitude spatiale d'une étude de cas basée sur un modèle économique d'équilibre partiel. Trois objectifs spécifiques sont dérivés : (1) réaliser une étude de cas C-LCA de scénarios de transport alternatifs pour étudier les bénéfices de la mise en œuvre d'une politique publique de transition énergétique en France d'ici 2050 avec une analyse d'incertitude pour explorer la force de nos conclusions, (2) réaliser des analyses de sensibilité globales pour identifier et quantifier les principales sources d'incertitude spatiale entre le modèle d'inventaire de premier plan issu de la modélisation économique d'équilibre partiel, le modèle d'inventaire de fond et les facteurs de caractérisation, (3) proposer une stratégie pour réduire l'incertitude spatiale pour notre étude de cas C-LCA en priorisant la régionalisation.Résultats et discussionLes résultats montrent que la mise en œuvre de scénarios de transport alternatifs en conformité avec la politique publique de transition énergétique en France est bénéfique pour certaines catégories d'impact (CI) (réchauffement climatique, acidification marine, eutrophisation marine, acidification terrestre, eau thermiquement polluée, formation d'oxydants photochimiques et formation de particules), avec un niveau de confiance de 95%. Pour les autres CI, une réduction de l'incertitude est nécessaire pour déterminer les conclusions avec un niveau de confiance similaire. Les variables d'entrée avec une variabilité spatiale provenant du modèle économique d'équilibre partiel contribuent de façon significative à l'incertitude spatiale de l'ACV-C et devraient être priorisées pour la réduction de l'incertitude spatiale. En outre, les facteurs de caractérisation contribuent de manière significative aux résultats de l'incertitude spatiale pour tous les CI régionalisés (sauf le CI d'occupation des sols).ConclusionsIl convient d'explorer les moyens de réduire l'incertitude spatiale de la modélisation économique. La réduction de l'incertitude afin d'améliorer la phase d'interprétation et la prise de décision doit être priorisée en fonction de l'objectif et de la portée de l'étude ACV. En outre, l'utilisation de FC régionalisées dans l'ACV-C semble être pertinente, et les outils de calcul de l'ACV-C devraient être adaptés en conséquence.

Background Recent advances in sequencing technologies have led to an explosion in the number of genomes available, but accurate genome annotation remains a major challenge. The prediction of protein-coding genes in eukaryotic genomes is especially problematic, due to their complex exon–intron structures. Even the best eukaryotic gene prediction algorithms can make serious errors that will significantly affect subsequent analyses. Results We first investigated the prevalence of gene prediction errors in a large set of 176,478 proteins from ten primate proteomes available in public databases. Using the well-studied human proteins as a reference, a total of 82,305 potential errors were detected, including 44,001 deletions, 27,289 insertions and 11,015 mismatched segments where part of the correct protein sequence is replaced with an alternative erroneous sequence. We then focused on the mismatched sequence errors that cause particular problems for downstream applications. A detailed characterization allowed us to identify the potential causes for the gene misprediction in approximately half (5446) of these cases. As a proof-of-concept, we also developed a simple method which allowed us to propose improved sequences for 603 primate proteins. Conclusions Gene prediction errors in primate proteomes affect up to 50% of the sequences. Major causes of errors include undetermined genome regions, genome sequencing or assembly issues, and limitations in the models used to represent gene exon–intron structures. Nevertheless, existing genome sequences can still be exploited to improve protein sequence quality. Perspectives of the work include the characterization of other types of gene prediction errors, as well as the development of a more comprehensive algorithm for protein sequence error correction.

In many real-world environments, only partial observations are provided, thus presenting challenges for Anticipatory Learning Classifier Systems (ALCS). The perceptual aliasing issue occurs when systems cannot differentiate situations that are truly distinct. To tackle the perceptual aliasing issue, ALCS classifiers can be chained in order to build Behavioral Sequences. Those sequences permit ALCS to deal with this issue, but they have never been implemented within ACS2 (Anticipatory Classifier System 2), although this is one of the most advanced ALCS. This paper introduces a novel learning classifier system, BACS, that integrates Behavioral sequences to ACS2. This integration required the adaptation of the action selection policy, the integration of an aliasing detection algorithm that let the system build behavioral classifiers, and the adaptation of the anticipatory learning process. The results obtained over a maze environment benchmark show that behavioral sequences are a promising approach to address the perceptual aliasing issue.

In this paper, we use a real valued genetic algorithm (GA) to model a large noisy periodic signal. The information that must be extracted are the amplitude, angular velocity and phase of the sines composing the signal. The algorithm outperforms the Fourier Transform (FT) method which has limitations when it comes to determining the phase component of a real part signal. The GA returns the phase component of the signal without the need of de-noising in a very short time, thanks to an efficient parallelization on GPGPU cards (≈ X400 acceleration compared to the same sequential code).

This paper presents a novel multi-threaded quantum inspired optimization algorithm targeted at global search in continuous domains. The proposed approach is based on a Diffusion Monte Carlo (DMC) physical model and is characterized by a set of parallel quantum walk processes. The effectiveness of the proposed algorithm is demonstrated by experimental results on the 24 noiseless functions from the Black Box Optimization Benchmark of the Comparing Continuous Optimization benchmarking platform (COCO).

We study a class of multi-species birth-and-death processes going almost surely to extinction and admitting a unique quasi-stationary distribution (qsd for short). When rescaled by K and in the limit K ! +1,the realizations of such processes get close, in any fixed finite-time window, to the trajectories of a dynamical system whose vector field is defined by the birth and death rates. Assuming that this dynamical has a unique attracting fixed point, we analyzed in a previous work what happens for large but finite K, especially the different time scales showing up. In the present work, we are mainly interested in the following question: Observing a realization of the process, can we determine the so-called engineering resilience? To answer this question, we establish two relations which intermingle the resilience, which is a macroscopic quantity defined for the dynamical system, and the fluctuations of the process, which are microscopic quantities. Analogous relations are well known in nonequilibrium statistical mechanics. To exploit these relations, we need to introduce several estimators which we control for times between log K (time scale to converge to the qsd) and $\exp(K)$ (time scale of mean time to extinction)., Nous étudions une classe de processus de naissance et mort avec plusieurs espèces dans la situation où l’extinction est certaine et la distribution quasi-stationnaire est unique. Si on fixe un intervalle de temps fini et qu’on normalise les réalisations d’un tel processus par un paramètre d’échelle K, elles deviennent arbitrairement proches, dans la limite K→+∞, des trajectoires d’un certain système dynamique dont le champ de vecteurs est défini à partir des taux de naissance et mort. Quand le système dynamique admet un seul point fixe attractif, nous avons précédemment analysé le comportement du processus pour des valeurs de K finies et pour des temps finis, c’est-à-dire le comportement intermédiaire entre les deux comportements limites évoqués ci-dessus. La question principale qui nous intéresse est la suivante : si on observe une réalisation du processus, pouvons-nous estimer la résilience au sens de l’ingénieur (engineering resilience) ? Pour répondre à cette question, nous démontrons deux relations entremêlant la résilience, qui est une quantité macroscopique définie pour le système dynamique sous-jacent, et les fluctuations du processus, qui sont, elles, des quantités microscopiques. De tels genres de relations sont bien connus en mécanique statistique hors d’équilibre. Afin d’exploiter ces relations nous introduisons plusieurs estimateurs empiriques que nous parvenons à contrôler pour des temps entre logK, qui est l’échelle de temps pour observer la convergence vers la distribution quasi-stationnaire, et exp(K), qui est l’échelle du temps moyen d’extinction.

Quantum-inspired algorithms are efficient for solving global search optimization problems. Nevertheless, their application is limited by two main requirements: a knowledge of a cost function and a big computational effort. To address both limitations, this paper presents a global optimization algorithm mixing a Quantum Diffusion Monte Carlo (DMC) method and an Evolution Strategy (ES). The proposed approach takes advantage of quantum models for efficiently finding solutions close the global optimum by applying the DMC. However, the DMC needs to be adapted to an optimization process. In order to improve relatively slow convergence rate of the DMC, we integrate a step size control for a diffusion displacement process of the DMC. Then, an (1+1)-ES is applied to find the minimum value of the cost function. Experimentation shows that the proposed approach can find global minima with fewer objective function evaluations than Quantum Particle Swarm Optimization and Quantum Annealing. Thus, it seems to be promising for solving black-box problems, where a reliable trade-off between exploration and exploitation is required.

This paper proposes a fast evolutionary algorithm for large-scale multi-objective optimization problems (MOPs), which widely exist in real-world applications [3, 6]. Many well-established multi-objective evolutionary algorithms (MOEAs) can not ensure necessary Runtime (RT) and values of performance metrics (Hypervolume (HV), Inverted Generational Distance (IGD)) for such kind of MOPs. The proposed archive-based algorithm provides better values of mentioned metrics due to its low complexity, simplified architecture and efficiency of genetic operators. Experimental results on three-objective and on two-objective benchmark suites (DTLZ [15], COCO 2018 Blackbox Optimization Benchmark (BBOB-biobj) [8]) demonstrate superiority of suggested algorithm in terms of performance metrics values and of RT over referenced MOEAs.

In life cycle assessment (LCA), temporal considerations are usually lost during the life cycle inventory calculation, resulting in an aggregated “snapshot” of potential impacts. Disregarding such temporal considerations has previously been underlined as an important source of uncertainty, but a growing number of approaches have been developed to tackle this issue. Nevertheless, their adoption by LCA practitioners is still uncommon, which raises concerns about the representativeness of current LCA results. Furthermore, a lack of consistency can be observed in the used terms for discussions on temporal considerations. The purpose of this review is thus to search for common ground and to identify the current implementation challenges while also proposing development pathways.This paper introduces a glossary of the most frequently used terms related to temporal considerations in LCA to build a common understanding of key concepts and to facilitate discussions. A review is also performed on current solutions for temporal considerations in different LCA phases (goal and scope definition, life cycle inventory analysis and life cycle impact assessment), analysing each temporal consideration for its relevant conceptual developments in LCA and its level of operationalisation.We then present a potential stepwise approach and development pathways to address the current challenges of implementation for dynamic LCA (DLCA). Three key focal areas for integrating temporal considerations within the LCA framework are discussed: i) define the temporal scope over which temporal distributions of emissions are occurring, ii) use calendar-specific information to model systems and associated impacts, and iii) select the appropriate level of temporal resolution to describe the variations of flows and characterisation factors.Addressing more temporal considerations within a DLCA framework is expected to reduce uncertainties and increase the representativeness of results, but possible trade-offs between additional data collection efforts and the increased value of results from DLCAs should be kept in mind.

International audience; Ongoing debates focus on the role of forest-sourced bioenergy within climate mitigation efforts, due to the long rotation lengths of forest biomass. Valuing sequestration is debated due to its reversibility; however, dynamic modelling of biogenic carbon (Cbio) flows captures both negative and positive emissions. The objective of this work is to respond to the key issue of timing sequestration associated with two opposed modelling choices (historic vs. future) in the context of dynamic life cycle assessment (LCA). Methods: The outputs of a partial-equilibrium model are used to inform prospective evaluations of the use of forest wood residues in response to an energy transition policy. Dynamic forest carbon modelling represents the carbon cycle between the atmosphere and technosphere: Cbio fixation and release through combustion and/or decay. Time-dependent characterization is used to assess the time-sensitive climate change effects. The two Cbio sequestration perspectives for bioenergy (forest biomass use) and reference (no use) scenarios are contrasted to assess (i) their temporal profiles, (ii) their climatic consequences concerning C-complete (fossil + biogenic C) vs. C-neutral (fossil C) approaches, and (iii) the implications of comparing the two approaches with dynamic LCA. Results and discussion: Full lifetime carbon accounting confirms that Cbio entering the bioenergy system equals the Cbio leaving it in the net balance, but not within annual dynamic balances, which alter the atmospheric greenhouse gas composition. The impacts of the historic approach differed considerably from those of the future. Moreover, the 'no use' scenario yielded higher forcing effects than the 'bioenergy' due to the higher methane proportions. The chicken-egg dilemma arises in attributional LCA: as the forcing depends on the timing of the Cbio sequestration and its allocation to a harvest activity. A decision tree supported by case study applications provides general rules for selecting the adequate time-related modelling approach based the criteria of provision of wood and regrowth from managed and unmanaged forests, determined by the origin of biotic resources and related spheres. Conclusions: Excluding dynamic Cbio introduces under- (future) or over- (historic) estimation of the results, misleading mitigation decisions. Further research is needed to close the gap between forest stand and landscape level, addressing issues beyond the chicken-egg dilemma and developing complete dynamic LCA studies.

This papers introduces BACS, a learning classifier system that integrates Behavioral Sequences to ACS2 (Anticipatory Classifier System 2), in order to address the Perceptual Aliasing Issue: this issue occurs when systems can not differentiate situations that are truly distinct in partially observable environments. In order to permit this integration, BACS implements (1) an aliased-state detection algorithm allowing the system to build behavioral classifiers and (2) an evolved Anticipatory Learning Process. A study of the capabilities of BACS is presented through a thorough benchmarking on 23 mazes. The obtained results show that Behavioral Sequences are a suitable approach to address the perceptual aliasing issue.

International audience; PurposeRegionalization in life cycle assessment (LCA) aims to increase the representativeness of LCA results and reduce the uncertainty due to spatial variability. It may refer to adapting processes to better account for regional technological specificities (inventory regionalization) or adding of spatial information to the elementary flows (inventory spatialization) which allow using more regionalized characterization factors. However, developing and integrating regionalization requires additional efforts for LCA practitioners and database developers that must be prioritized.MethodsWe propose a stepwise methodology for LCA practitioners to prioritize data collection for regionalization based on global sensitivity analysis (GSA) using Sobol indices. It involves several GSA to select the impact categories (ICs) that require further inventory data collection (IC ranking), prioritize between inventory regionalization and inventory spatialization (LCA phase ranking), and target specific data to collect. Then we propose a method to derive sector-specific recommendations using statistical tests to prioritize inventory regionalization versus spatialization and the ICs on which to focus inventory data collection. These recommendations are meant to help LCA practitioners and database developers define their strategy for regional data collection by focusing on data that have the highest potential to reduce the uncertainty of the results.Results and discussionThe applicability of the methodology is illustrated through three case studies using the ecoinvent v3 database and the regionalized impact methodology IMPACT World+: one on prioritizing data collection in a single biofuel product system and two meta-analyses of all product systems in two distinct economic sectors (biofuel production and land passenger transport). Recommendations for regionalization can be derived for an economic sector and appear to be different from one economic sector to another. GSA seems to be more relevant to prioritize regionalization efforts than an impact contribution analysis (ICA) approach often used to prioritize data collection in LCA. However, further improvements, such as accounting for spatial correlations and better computational times for GSA, are required to implement it in LCA.ConclusionsWe recommend using the methodology based on GSA to efficiently prioritize regionalization efforts between ICs and between inventory regionalization and inventory spatialization. We proved that the implementation of IC ranking and LCA phase ranking is computationally feasible and therefore invite current LCA software providers to unlock this new horizon in LCA interpretation. We also invite to expand the meta-analysis to all sectors in an LCA database.

We consider a class of birth-and-death processes describing a population made of $d$ sub-populations of different types which interact with one another. The state space is $\mathbb{Z}^d_+$ (unbounded). We assume that the population goes almost surely to extinction, so that the unique stationary distribution is the Dirac measure at the origin. These processes are parametrized by a scaling parameter $K$ which can be thought as the order of magnitude of the total size of the population at time $0$. For any fixed finite time span, it is well-known that such processes, when renormalized by $K$, are close, in the limit $K\to+\infty$, to the solutions of a certain differential equation in $\mathbb{R}^d_+$ whose vector field is determined by the birth and death rates.We consider the case where there is a unique attractive fixed point (off the boundary of the positive orthant) for the vector field (while the origin is repulsive). What is expected is that, for $K$ large, the process will stay in the vicinity of the fixed point for a verylong time before being absorbed at the origin. To precisely describe this behavior, we prove the existence of a quasi-stationary distribution (qsd, for short). In fact, we establish a bound for the total variation distance between the process conditioned to non-extinction before time $t$ and the qsd. This bound is exponentially small in $t$, for $t\gg \log K$. As a by-product, we obtainan estimate for the mean time to extinction in the qsd. We also quantify how close is the law of the process (not conditioned to non-extinction) either to the Dirac measure at the origin or to the qsd, for times much larger than $\log K$ and much smaller than the mean time to extinction, which is exponentially large as a function of $K$. Let us stress that we are interested in what happens for finite $K$. We obtain results much beyond what large deviation techniques could provide., Nous considérons une classe de processus de naissance-et-mort décrivant une population constituée de d sous-populations de types différents qui interagissent entre elles. L'espace d'état est $\mathbb{Z}^d_+$ (il est donc non borné). Nous supposons que la population s'éteint presque sûrement, de sorte que l'unique distribution de probabilité stationnaire est la masse de Dirac à l'origine. Nous faisons dépendre ces processus d'un paramètre d'échelle K qu'on peut interpréter comme l'ordre de grandeur de la taille totale de la population au temps 0. Etant donné un intervalle de temps, il est bien connu que de tels processus, normalisés par K, sont proches, dans la limite K → +∞, des solutions d'une certaine équation différentielle dans $\mathbb{R}^d_+$ dont le champ de vecteurs est déterminé par les taux de naissance et de mort du processus. Nous considérons le cas où le champ de vecteurs possède un unique point fixe attractif à l'intérieur de l'orthant positif, tandis que l'origine est un point fixe répulsif. On s'attend à ce que, pour K grand, le processus reste dans le voisinage du point fixe attractif pendant très longtemps avant d'être absorbé à l'origine. Afin de décrire précisément ce comportement, nous démontrons l'existence dune distribution quasi-stationnaire (dqs, en abrégé). Nous établissons une borne pour la distance en variation totale entre le processus conditionné à ne pas s'éteindre avant le temps t et la dqs. Cette borne est exponentiellement petite en t pour t log K. En particulier, nous obtenons une estimation du temps moyen d'extinction dans la dqs. Nous quantifions également la distance entre le processus (non conditionné à la non-extinction) et une certaine combinaison convexe de la masse de Dirac à l'origine et de la dqs, ceci pour des temps beaucoup plus grands que log K et beaucoup plus petits que le temps moyen d'extinction, qui est exponentiellement grand en K. Nous attirons l'attention sur le fait que nous sommes intéressés par ce *

To account for the possibility of an externally driven taxis in active systems, we develop a model of a guided active drift which relies on the presence of an external guiding field and a vectorial coupling between the mechanical degrees of freedom and a chemical reaction. To characterize the ability of guided active particles to carry cargo, we generalize the notion of Stokes efficiency extending it to the case of stall conditions. To show the generality of the proposed mechanism, we discuss guided electric circuits capable of turning fluctuations into a directed current without a source of voltage.

International audience; To decrease the use of fossil fuels and face the energetic demand, the integration of renewable energy is a necessary step. Part of this renewable energy can be supplied by the production of electricity from photovoltaic panels and windfarms. The massive use of these intermittent energies will lead to overproduction periods, and there is consequently a need to convert this surplus of electricity into a storable form of energy. Power-togas (PtG) technology consists in using electricity to convert water into hydrogen by electrolysis, and then to synthetize methane from carbon dioxide and hydrogen. Techno-economic and Life Cycle Assessment of methane production via the combination of anaerobic digestion and PtG technology have been applied to sewage sludge valorization. Process studies and equipment design have been addressed considering already available technologies. Sensitivity analyses have been done on biogas upgrading technologies, electricity prices, annual operation time and composition of the electricity mix with also a comparison between PtG and direct injection. It appears that the more the electricity is expensive, the longer the operation time of the methanation process must be to be competitive with injection of methane from biogas. Reduction of electricity consumption of the electrolysis step decreases production costs. Even if the current context does not feature adapted conditions to ensure an economically viable chain, the evolution of the energetic context in the next few years as well as the expected technological improvements will contribute to overall cost reduction. From an environmental point of view, continuous PtG generates more greenhouse gases than direct injection, but intermittent operation with use of renewable electricity can significantly reduce GHG emissions. From an endpoint impacts perspective, impact from continuous PtG are higher than biogas upgrading, but much lower than fossil energy. Future development of low electricity consumption of the electrolysis process, and integration of renewable credits from CO 2 valorization can increase the competitiveness of this technology.

Detecting anomalies in time series in real time can be challenging, in particular when anomalies can manifest themselves at different time scales and need to be detected with minimal latency. The need for lightweight real-time algorithms has risen in the context of Cloud computing, where thousands of devices are monitored and deviations from normal behaviour must be detected to prevent incidents. However, this need has yet to be addressed in a way that actually scales to the size of today’s network infrastructures. Typically, time series generated by human activity often exhibit daily and weekly patterns creating long-term dependencies that are difficult to process. In such cases, the euclidean distance between subsequences of the time series, or euclidean anomaly score, can be a very effective tool to achieve good detection within constrained latency; however, this computation has a quadratic complexity and a computational footprint too high for any realistic application. In this paper, we propose SCHEDA (Sampled Causal Heuristics for Euclidean Distance Approximation), a collection of three heuristics designed to approximate the euclidean anomaly score with a low computational footprint in time series with long-term dependencies. Our design goals are a low computational cost, the possibility of real-time operation and the absence of tuning parameters. We benchmark SCHEDA against ARIMA and the euclidean distance and show that in typical monitoring scenarios, it outperforms both at only a fraction of the computational cost.

International audience

Conventional wastewater treatment (WWT) is currently based on the action of heterotrophic and nitrifying bacteria supported by mechanical aeration in a sequence of interconnected aerobic, anoxic, and anaerobic processes. It provided satisfactory levels of pollutant removal, but at the expenses of high-energy consumption and environmental impacts (high CO2 footprint, nutrient losses, and secondary pollution due to the use of chemicals).

Despite the rich research efforts and the expressive power of existing Artificial Immune System models, the community is struggling to devise operational tools that go beyond ‘immunity marketing’. We claim that this matter of fact lays in the inadequacy between the analysis metaphors used for AIS algorithms and the actual requirements of distributed systems that we want to see ‘immunized’. We strongly believe that building efficient immune systems require going back to the source of immune model analysis: the work of Francesco Varela on cognitive sciences and the need for a shift of cognitive analysis towards enactement,.e. the analysis of system cognition as the control of the very interactions of the system with its environment. This radical shift back to the origin of artificial immunity requires to extract the properties of interest of such an enacting system, and to characterize its implementation. Such a ‘general, reusable solution to a commonly occurring problem within a given context in software design’ is a well-known artefact in software engineering: a Design Pattern. Artificial Immune Systems are thus a Design Pattern for enacting software, rather than a family of bio-inspired algorithms as we usually consider.

The problem of finding the optimal incident wave of given duration that maximizes the efficiency of conversion of wave energy into work in percussive drilling with detachable drill bit is considered. The drill rod is modelled as 1D linearly elastic and the drill bit as a rigid mass. The bit/rock interaction is described by a history-dependent force versus penetration relation with different constant slopes for primary loading and unloading/reloading. A functional expressing the dependence of the efficiency on the shape of an arbitrary incident wave of given duration is derived and maximized. For short incident waves, there is a weak influence of the bit mass on the optimal wave shape which is nearly rectangular. For longer incident waves, there is a strong influence of the bit mass on the optimal wave shape which significantly differs from rectangular. The efficiencies for optimal waves approach those for rectangular waves for short waves. For long waves they approach or assume values which are independent of wave duration but decrease with increasing bit mass. Relative to commonly-used rectangular waves significant increase in efficiency can be achieved through optimization of the wave shape if the wave is not too short. Optimal incident waves can be realized accurately, e.g., by piezoelectric means.

The evolution of cellular automata and of finitely anticipative transformations is studied by using right sets. These are the sets of symbols that are compatible with a past of a position and the respective coordinate of the transformation. Our main result shows, under some suitable conditions, that if the entropy converges to zero then the right sets increase towards the whole alphabet. We discuss these concepts with Wolfram automata.