Heterogeneous Parameters Research Articles

Assessment of the interactions between soil–biosphere–atmosphere (ISBA) land surface model soil hydrology, using four closed-form soil water relationships and several lysimeters

Abstract. Soil water drainage is the main source of groundwater recharge and river flow. It is therefore a key process for water resource management. In this study, we evaluate the soil hydrology and the soil water drainage, simulated by the interactions between soil–biosphere–atmosphere (ISBA) land surface model currently used for hydrological applications from the watershed scale to the global scale, where parameters are generally not calibrated. This evaluation is done using seven lysimeters from two long-term model approach sites measuring hourly water dynamics between 2009 and 2019 in northeastern France. These 2 m depth lysimeters are filled with different soil types and are either maintained as bare soil or covered with vegetation. Four closed-form equations describing soil water retention and hydraulic conductivity functions are tested, namely the commonly used equations from Brooks and Corey (1966) and van Genuchten (1980), a combination of the van Genuchten (1980) soil water retention function with the Brooks and Corey (1966) unsaturated hydraulic conductivity function, and, for the very first time in a land surface model (LSM), a modified version of the van Genuchten (1980) equations, with a new hydraulic conductivity curve proposed by Iden et al. (2015). The results indicate good performance by ISBA with the different closure equations in terms of soil volumetric water content and water mass. The drained flow at the bottom of the lysimeter is well simulated, using Brooks and Corey (1966), while some weaknesses appear with van Genuchten (1980) due to the abrupt shape near the saturation of its hydraulic conductivity function. The mixed form or the new van Genuchten (1980) hydraulic conductivity function from Iden et al. (2015) allows the solving of this problem and even improves the simulation of the drainage dynamic, especially for intense drainage events. The study also highlights the importance of the vertical heterogeneity of the soil hydrodynamic parameters to correctly simulate the drainage dynamic, in addition to the primary influence of the parameters characterizing the shape of the soil water retention function.

Bayesian regularization networks for micropolar ternary hybrid nanofluid flow of blood with homogeneous and heterogeneous reactions: Entropy generation optimization

This study aims to analyze a Bayesian regularization backpropagation algorithm for micropolar ternary hybrid nanofluid flow over curved surfaces with homogeneous and heterogeneous reactions, Joule heating and viscous dissipation. The ternary hybrid nanofluid consists of nanoparticles of titanium oxide (TiO2), copper oxide (CuO), and silicon oxide (SiO2), with blood as the base fluid. The governing partial differential equations for the fluid flow are converted into ordinary differential equations using a group of self-similar transformations. The ordinary differential equations are solved using an appropriate shooting algorithm in MATLAB. The effects of physical parameters including curvature, micro-polar, radiation, magnetic, Prandtl, Eckert, Schmidt, and homogeneous and heterogeneous chemical reaction parameters are analyzed for velocity, micro rotational, temperature, and concentration profile. Physical quantities of engineering interest like heat transfer rate, mass transfer rate, skin friction coefficient, couple stress coefficient, and entropy generation are also discussed in this study. A Bayesian regularization backpropagation algorithm is also designed for the solution of the ordinary differential equations. The obtained network is analyzed using training state, performance, error histograms, model response, Error autocorrelation, and input-error correlation plots. It is observed that the entropy generation and the Bejan number increase for enhancing Brinkman and radiation parameter. Clinical researchers and biologists may use the results of this computational study to forecast endothelial cell damage and plaque deposition in curved arteries, by which the severity of these conditions can be reduced.

Derivation of heterogeneous material distributions and their sensitivity to HM-coupled two-phase flow models exemplified with the LASGIT experiment

Advective gas transport in bentonite, a possible buffer material in repositories for radioactive materials, is difficult to simulate in numerical continuum models, partly due to the complicated microstructure of bentonite. To generate reliable models of repositories nevertheless, spatially distributed heterogeneous material properties can be used to allow localization of gas flow. In this study, a pore-size-dependent stochastic approach of the gas entry pressure is derived from Mercury Intrusion Porosimetry, which is used to replicate measurements from the LASGIT experiment. In addition, three benchmark tests are simulated to investigate the dependence of heterogeneous distributions of material properties on the mesh discretization, the temporal dependence, and the coupling between the processes influenced by the heterogeneous parameters. The numerical modeling results of the LASGIT experiment show that the onset of gas flow into the system and the subsequent increase in pressure and stress can be well reproduced using heterogeneous distributions. Compared to a model with homogeneous material properties, heterogeneous distributions may allow the generation of dilatancy-controlled microfractures—an important feature with regard to the advective gas flow in bentonites. However, it can be observed that the heterogeneous distributions in LASGIT are less significant, as technical gaps or differences in material types could have a greater impact.

Hierarchical control strategy of thermostatically controlled load considering multiple factors

In this article, a distributed hierarchical control strategy of thermostatically controlled loads (TCLs) is proposed, which can effectively utilize TCLs to provide power dispatching service in a microgrid composed of buildings. Firstly, the distributed consensus protocol of TCLs is studied, and the state space of a TCL is established. The optimal control performance is achieved by selecting coupling strength and control gain. Then, the dispatching power of each building is reasonably distributed through pinning control, and consensus coefficient and proportional coefficient are determined according to the heterogeneity of building parameters. Finally, the case studies were conducted to validate the proposed novel distributed hierarchical control scheme, which can provide effective power dispatching services and realize the fair distribution of the dispatching requirements among all TCLs. The proposed strategy also takes into account the multiple factors of each TCL such as heterogeneous parameters, communication burden and user's actual conditions, which improves the feasibility and expansibility of the dispatching process.

Definition of learning curve for thyroidectomy: systematic review on the different approaches.

Thyroidectomy is one of the most common surgical procedures carried out worldwide and it has evolved in recent years with alternative approaches. With the advent of minimally invasive techniques, the learning curve (LC) concept has become a fundamental "dogma". A literature search, according to the PRISMA guidelines, was performed via PubMed (MEDLINE), Scopus, Cochrane Library, EMBASE, and Web of Science. Only studies assessing the learning process to thyroidectomy (including hemi- and total thyroidectomy), reporting a minimum of 30 procedures and describing clearly the minimum number of performances required to achieve proficiency and the main evaluation items used to establish it, were included. Conventional, endoscopic and robotic approaches were separately analyzed. Only English-language studies were considered. Forty-five relevant studies were selected for the analysis [respectively 16 concerning robotic thyroidectomy (RT), 22 endoscopic thyroidectomy (ET), 6 mini-invasive video assisted thyroidectomy (MIVAT), 1 conventional thyroidectomy (CT)]. The number of procedures required for a single surgeon to achieve competence and the parameters used to define surgical proficiency were fully investigated for each individual technique. Our research shows how the current literature lacks an objective definition of the LC concept. The heterogeneity of analysis methodologies and parameters evaluated, the various surgical techniques and training background of single surgeons, make it impossible to draw univocal results. Future studies should consider confounding factors and establish criteria that should be consensually recognized in the assessment of surgical performances and skills.

Investigation and Analysis on Energy Efficient Stable Election Protocol (SEP) for WSN

Abstract: We study the impact of heterogeneity of nodes, in terms of their energy, in wireless sensor networks that are hierarchically clustered. In these networks some of the nodes become cluster heads, aggregate the data of their cluster members and transmit it to the sink. We assume that a percentage of the population of sensor nodes is equipped with additional energy resources—this is a source of heterogeneity which may result from the initial setting or as the operation of the network evolves. We propose SEP, a heterogeneous-aware protocol to prolong the time interval before the death of the first node (we refer to as stability period), which is crucial for many applications where the feedback from the sensor network must be reliable. SEP is based on weighted election probabilities of each node to become cluster head according to the remaining energy in each node. We show by simulation that SEP always prolongs the stability period compared to (and that the average throughput is greater than) the one obtained using current clustering protocols. We conclude by studying the sensitivity of our SEP protocol to heterogeneity parameters capturing energy imbalance in the network. We found that SEP yields longer stability region for higher values of extra energy brought by more powerful nodes. Sensor nodes usually have limited energy supply and they are impractical to recharge. How to balance traffic load in sensors in order to increase network lifetime is a very challenging research issue. Many clustering algorithms have been proposed recently for wireless sensor networks (WSNs). The use of mobile sinks has been shown to be an effective technique to enhance network performance features such as latency, energy efficiency, network lifetime, etc. In this paper, a modified Stable Election Protocol (SEP), which employs a mobile sink, has been proposed for WSNs with non-uniform node distribution. The decision of selecting cluster heads by the sink is based on the minimization of the associated additional energy and residual energy at each node. Besides, the cluster head selects the shortest path to reach the sink between the direct approach and the indirect approach with the use of the nearest cluster head. Simulation results demonstrate that our algorithm has better performance than traditional routing algorithms, such as LEACH and SEP.

Analysis on mathematical model of convection system of micropolar fluid as darcy forchheimer flow undergoes heterogeneous and homogeneous chemical reaction

In this article, to study the fluid behavior on various specific conditions to improve the heat transfer system and here, analysis on mathematical model of dynamic fluid consist of micropolar fluid has allowed micro rotational effect with laminar flow of Darcy forchheimer model which allow inertia effect has incorporated with heterogeneous and homogenous chemical reaction undergoes heat exchanger system with boundary layer problem has modeled. The mathematical model of fluid mechanic governing equations are in the form of partial differential equation (PDE) and similarity transformation into numerical methods (PC4-FDM) of predictor and corrector technique undergoes discretized mesh point and convergence with fourth order finite difference method and shooting method is also equipped to get better solution. The additions of significant heterogeneous parameter depicts that increasing behavior in fluid concentration and homogeneous parameter depicts that decreasing in fluid concentration by allowed micro rotations leads to collision and on increasing the Eckert number related to viscous dissipation has exhibited that increased fluid velocity and decreased fluid temperature. Micro rotation parameter exhibits that similar increased fluid velocity and slight decreased in temperature of the fluid. Darcy forchheimer parameter which related to inertial effect has depicts that decreased in velocity with increased temperature of the fluid in convection system. Due to Industrialization, the study of convection heat transfer system has enormous scope which has necessity to improving the heating and cooling system of industrial mass machineries like powerplant, waste heat recovery unit, pharmaceutical industries etc.

Efficient simulation of CO2 migration dynamics in deep saline aquifers using a multi-task deep learning technique with consistency

CO2 sequestration and storage in deep saline aquifers is a promising technology for mitigating the excessive concentration of the greenhouse gas in the atmosphere. However, accurately predicting the migration of CO2 plumes requires complex multi-physics-based numerical simulation approaches, which are prohibitively expensive due to highly nonlinear coupled governing equations and uncertainties in heterogeneous spatial parameter distributions. To address this challenge, we developed an end-to-end deep learning workflow employing encoder–decoder architectures with residual network (ResNet) to efficiently predicts the spatial–temporal evolution of the solution CO2-brine ratio (Rs) and gas saturation (Sg) – the two essential tasks for quantifying the amount of trapped CO2 – given heterogeneous permeability fields as input. Specifically, we introduce a general multi-task learning with consistency (MTLC) framework to simultaneously predict Rs and Sg. The MTLC model leverages related tasks with less computational expensive labeled datasets to improve generalization ability. In our study, predictions for multiple tasks from the same permeability realization are not independent but expected to be consistent, as the proposed framework utilizes data-driven cross-task consistency constraints to augment learning of related tasks. Our deep learning model is trained based on physical trapping mechanisms, which play a dominant role in the CO2 migration process. The results demonstrate that the MTLC model with joint learning yields more accurate predictions and improved generalization for predicting CO2 migration in several test cases. Furthermore, our workflow is 105 times faster than a high-fidelity physics-based numerical simulator, making it a viable alternative for field-scale applications.

Clustering heterogeneous financial networks

AbstractWe develop a convex‐optimization clustering algorithm for heterogeneous financial networks, in the presence of arbitrary or even adversarial outliers. In the stochastic block model with heterogeneity parameters, we penalize nodes whose degree exhibit unusual behavior beyond inlier heterogeneity. We prove that under mild conditions, this method achieves exact recovery of the underlying clusters. In absence of any assumption on outliers, they are shown not to hinder the clustering of the inliers. We test the performance of the algorithm on semi‐synthetic heterogenous networks reconstructed to match aggregate data on the Korean financial sector. Our method allows for recovery of sub‐sectors with significantly lower error rates compared to existing algorithms. For overlapping portfolio networks, we uncover a clustering structure supporting diversification effects in investment management.

The effect of dapagliflozin therapy on ventricular repolarization parameters in electrocardiography in patients with diabetic cardiovascular disease

BackgroundDapagliflozin is an agent that has both antihyperglycemic effects and is significantly associated with a lower risk of cardiovascular events in patients with diabetes mellitus (DM). However, there is insufficient data and information about the effect of dapagliflozin on electrocardiographic (ECG) parameters. AimThe effects of dapagliflozin on ventricular repolarization parameters have not been fully elucidated yet. This study aimed to investigate whether dapagliflozin has a positive effect on ventricular repolarization heterogeneity parameters in patients with type 2 DM. MethodWe retrospectively enrolled 140 patients with a known diagnosis of type 2 DM who were newly prescribed dapagliflozin in addition to standard anti-diabetic therapy. The patients were divided into two groups according to whether they had cardiovascular disease (CVD). The effect of dapagliflozin treatment on ventricular repolarization parameters (frontal plane QRST angle, Tp-e interval, Tp-e/QTc, QTc, and QTc dispersion) was investigated, patient groups were compared before and after treatment. ResultsAmong 140 patients, 70 (50 %) had CVD and 70 (50 %) did not have CVD. Dapagliflozin treatment resulted in significant reductions in ventricular repolarization parameters over the study period in the CVD group with diabetes. Mean fQRST angle, Tp-e interval, Tp-e/QTc, QTc, and QTc dispersion were significantly lower than baseline values at 6-month follow-up visits in the CVD group (61.61 ± 9.22° vs 52.55 ± 8.31°, 74.45 ± 16.06 msec vs 63.27 ± 13.99 msec, 0.19 ± 0.03 vs 0.16 ± 0.03, 384.12 ± 47.93 msec vs 356.15 ± 43.31 mesc, 55.28 ± 5.50 msec vs 48.08 ± 6.48 msec for all pairwise comparisons p < 0.001). ConclusionSimilar antihyperglycemic effects were found with dapagliflozin treatment in patients with and without CVD. However, significant reductions in ventricular repolarization parameters were observed only in patients with CVD.

On the Influence of Organizational and Technical Measures on the Cascade Evolvement of Accidents in Electric Power Systems with Heterogeneous Parameters

The aim of the study is to assess the effect of various organizational and technical measures on the evolvement of emergency cascade processes in electric power systems (EPS) with heterogeneous parameters and to determine the limits of the maximum permissible operation modes. The article presents the results of electrical operation mode calculations in the test configurations of 6, 35, 110, and 500 kV EPS containing 25, 36 and 40 nodes, which illustrate the influence of heterogeneity of the EPS parameters on the evolvement of cascade emergency processes. An approach to taking into account the heterogeneity of EPS parameters in planning and maintaining the EPS operation modes is presented. To enhance the EPS survivability and reduce the risk of cascade emergency processes to occur and evolve, distributed generation facilities of different capacities and FACTS devices were used in the EPS. In addition, power flows were redistributed by changing the loading of existing power plants with different electricity generation costs. The proposed approach was used to analyze the evolvement of a cascade-type system accident that occurred in the EPS of the Republic of Bashkortostan. By using the developed approach, the possibility of cascade emergency processes to occur and evolve can be evaluated in planning and maintaining the EPC operation modes, as well as to develop EPS configuration and operation mode adjustment measures aimed at preventing such emergencies from occurring.

#4399 ALGORITHM-MANAGED DOSING AND PHARMACIST-MANAGED DOSING OF ERYTHROPOIETIN STIMULATING AGENTS IN RENAL ANEMIA: A SYSTEMATIC REVIEW

Abstract Background and Aims In clinical practice, the treatment of renal anemia is challenging. The attainment of target levels for hemoglobin is often low, due to the high incidence of infections, hyporesponsiveness to erythropoietin stimulating agents (ESA), and suboptimal prescribing of ESA and iron. Several interventions to improve the treatment of renal anemia have been developed, two of them being algorithm-managed dosing and pharmacist-managed dosing of ESA. We performed a systematic review to identify and summarize these two types of interventions and to determine their effectiveness in improving the treatment of renal anemia. Method We followed the PRISMA guidelines for systematic reviews. Studies that explored the effect of algorithm-managed and pharmacist-managed dosing of ESA in adult patients with renal anemia were evaluated for inclusion. No restrictions were set on outcome parameters. All observational and interventional studies that included a control group and had a follow-up of at least six months, were eligible for inclusion. Only full-length articles were considered for inclusion. PubMed, Embase, Web of Science, and the Cochrane Library were searched from their inception through July 2022. All included studies were evaluated by two independent reviewers. The quality of studies was assessed by the Newcastle Ottawa Scale and the risk of bias was assessed by the ROBINS-I and RoB1 tools. Data were summarized and tabulated. Studies were grouped according to intervention type, study design, and risk of bias. The protocol of this study was registered in PROSPERO (International Prospective Register of Systematic Reviews) with ID CRD42021243678. This study was funded by the Franciscus Gasthuis and Vlietland hospital. Results A total of 120 articles were assessed for inclusion, and after screening, 16 articles were included with a total number of 3777 patients. Available evidence was scarce and generally of low to moderate quality; only two RCTs could be identified. All but one of the other studies were observational in nature. The risk of bias was serious in all but one study. Study follow-up was relatively short, varying between six and thirteen months. In six studies, ESA dosing was pharmacist-managed and in one study ESA dosing was algorithm-managed, in nine studies a combination was used. The quality of the intervention description was low to moderate and interventions generally were not reproducible. Although heterogeneity was substantial for outcome parameters, four types of outcome parameters could be determined: hemoglobin/hematocrit, ESA dose and expenditure, iron status, and iron dose. Quantitative synthesis of data was not possible due to the substantial heterogeneity in outcome parameters and the high risk of bias. Therefore, the effectiveness of algorithm-managed and pharmacist-managed dosing of ESA in renal anemia could only be qualitatively assessed for the four types of outcome parameters. In six of the fifteen studies that reported on hemoglobin or hematocrit, the percentage of patients within target levels was significantly higher for the intervention group, whereas in four studies no significant difference was found. Six of the ten studies that reported on ESA dose or ESA expenditure found a significant decrease in ESA dose or expenditure in the intervention group, whereas two studies reported no significant difference. In five of the seven studies that reported on iron status, a significantly higher iron status was found in the intervention group, whereas one non-inferiority study reported no significant difference. In three of the four studies that reported on the iron dose, no significant difference was found between the intervention and the control group, whereas in one study the iron dose was significantly higher in the intervention group. Conclusion Available evidence was scarce with a high risk of bias, and quantitative data synthesis was not possible. Therefore, no definite conclusions could be drawn on the effectiveness of algorithm-managed and pharmacist-managed dosing of ESA in renal anemia. Consequently, no recommendations on the implementation of either of the two interventions could be made.

Estimation of heterogeneous permeability using pressure derivative data through an inversion neural network inspired by the Fast Marching Method

Reservoir heterogeneity significantly impacts the fluid flow behavior in porous media. In subsurface communities including hydrocarbon or geothermal recovery, geological storage of CO2 or H2, and hydrology, transient pressure data are often used to infer the subsurface rock properties (e.g., permeability field) due to its ready availability and quick response. The accurate estimation of such properties is critical to accurately predict fluid flow in porous media.In this work, we propose a deep learning (DL) approach inspired by the Fast Marching Method (FMM), namely the Inversion Neural Network (INN), to inversely infer heterogeneous reservoir model parameters using transient pressure data. The forward model used to generate training data for the INN is established based on a semi-analytic asymptotic solution to the diffusivity equation using the diffusive time of flight (DTOF). The FMM solves the Eikonal equation for the DTOF and provides the partial derivative of pressure drop to the natural logarithm of time (hereafter pressure derivative). The pressure derivative data are then used to predict reservoir model parameters by the INN. In the homogeneous scenario, the INN architecture is a relatively simple fully connected neural network as a proof-of-concept to validate the feasibility, and it directly correlates the permeability value with the pressure derivative. In the heterogeneous scenario, as the heterogeneous permeability field is estimated based on sparse observational data of pressure derivatives, we adopt the convolutional neural network (CNN) to flexibly deal with the image-based properties, and leverage transfer learning to efficiently train a robust INN in the heterogeneous scenario.We first validated that INN can infer the homogeneous permeability fields through numerical experiments by testing root-mean-square-error (RMSE) around 1.086 md. Inspired by that, observational data with different sparsity are used to train convolutional INN and predict heterogeneous permeability fields. As the number of observational locations (nobs) increases from 3 × 3 to 48 × 48, the testing RMSE in heterogeneous scenarios decreases from 187.510 md to 18.080 md. Besides, we found that transfer learning significantly improves the predictive accuracy at low nobs, with relative error decreased by 17.25%. Finally, noisy pressure derivative data under different nobs are used to history match the heterogeneous reservoir model, and INN can infer the permeability field with a low error of 9.6% at nobs=7×7. Without an iterative procedure for parameter estimation, the INN demonstrates to perform permeability inversion with CPU time in the magnitude of 9.2×10−4 s on a single GPU (NVIDIA Quadro P2200). The FMM-inspired INN sets up the basis for the accurate characterization of reservoir model heterogeneity by inverting pressure derivative data with both decent predictive accuracy and computational efficiency.

Spatial patterns and species associations of earthworms in relation to micro-topography and insolation

Despite significant advancements in our understanding of the relationship between soil animal life and environmental gradients, as well as their impact on ecosystem functioning, there is a growing need for quantitative research that considers the spatial and temporal heterogeneity of soil biodiversity. Such research can help to address current knowledge gaps and support the development of legislation and conservation measures aimed at preserving soil biodiversity in the future. Earthworms, one of the most important group of soil animals, show patchy horizontal distribution in response to spatial heterogeneity in soil physico-chemical parameters, food availability and biological interactions at various scales.The aim of this study was to explore whether the spatial and temporal variability and co-occurrence probability of earthworm species and ecological categories (epigeic, endogeic and anecic) can be characterized by exploiting habitat micro-topography and insolation.Field research was carried out in spring and summer at seven locations using a regular grid of 49 sampling points with 5 m inter-sample distance over an altitudinal gradient. The results of Geary's C and Moran's I indices, as well as the semivariograms calculated for each species, season and location confirmed a patchy spatial distribution. The range of semivariograms varied from 3.2 m to 32.8 m in spring, and from 0.3 m to 32.8 m in summer. Overall (all locations and both seasons), the mean range was similar for adult (8.4 m) and juvenile (9.2 m) earthworms. In spring, epigeic species had a larger range than endogeic species. Abundances of juvenile earthworms were negatively correlated with microtopographic elevation in spring (at most locations) and in summer (at all locations). The relationship between adult earthworms and microtopographic elevation could not be interpreted generally within season. Additionally, correlations between earthworm abundance and average monthly insolation were weak. Pairwise analysis of species co-occurrence revealed a random pattern for most species. In the analysis of co-occurrence of ecological categories, endogeics and epigeics had more positive probabilities than other pairs. The obtained results could be useable in soil biodiversity assessments, or creation of monitoring protocols.

Breast cancer diagnosis and prognosis using a high b-value non-Gaussian continuous-time random-walk model

To compare the diagnostic performance of mono-exponential model-derived apparent diffusion coefficient (ADC), continuous-time random-walk (CTRW) model-derived Dm, α, β and their combinations in discriminating malignancy of breast lesions, and investigate the association between model-derived parameters and prognosis-related immunohistochemical indices. A total of 85 patients with breast lesions (51 malignant, 34 benign) were analysed in this retrospective study. Clinical characteristics include oestrogen receptor (ER), progesterone receptor (PR), human epidermal receptor 2 (HER2), and Ki-67. The ADC was fitted using a mono-exponential model (b-values=0, 800 s/mm2), while Dm, α, and β were fitted using a CTRW model. Independent Student's t-test and the Mann-Whitney U-test were used for the comparison of parameters. Discrimination performance was accomplished by receiver operating characteristic (ROC) analysis, and Spearman's correlation analysis was used to explore the association between immunohistochemical indices and diffusion parameters, the statistical significance level was p<0.05. Dm and ADC demonstrated similar performance in differentiating malignant and benign lesions (AUC=0.928 versus 0.930), while the combination of Dm, α, and β could improve the AUC to 0.969. The combined parameter generated by ADC, Dm, α, and β was effective in identifying the ER+/ER- and PR+/PR- patients. Temporal heterogeneity parameter α correlated significantly with the expression of PR. Diffusion parameters derived from the CTRW model could effectively discriminate the malignancy of breast lesions. Meanwhile, the hormone receptor expression could be distinguished by combined diffusion parameters, and have the potential to reflect the prognosis.

HeterPS: Distributed deep learning with reinforcement learning based scheduling in heterogeneous environments

Deep neural networks (DNNs) exploit many layers and a large number of parameters to achieve excellent performance. The training process of DNN models generally handles large-scale input data with many sparse features, which incurs high Input/Output (IO) cost, while some layers are compute-intensive. The training process generally exploits distributed computing resources to reduce training time. While heterogeneous computing resources, e.g., CPUs, GPUs of multiple types, are available for the distributed training process, the scheduling of multiple layers to diverse computing resources remains critical for the training process. To efficiently train a DNN model using the heterogeneous computing resources, we propose a distributed framework, i.e., Heterogeneous Parameter Server (HeterPS), composed of a distributed architecture and a Reinforcement Learning (RL)-based scheduling method. The advantages of HeterPS are three-fold compared with existing frameworks. First, HeterPS enables efficient training process of diverse workloads with heterogeneous computing resources. Second, HeterPS exploits an RL-based method to efficiently schedule the workload of each layer to appropriate computing resources to minimize the cost while satisfying throughput constraints. Third, HeterPS manages data storage and data communication among distributed computing resources. We carry out extensive experiments to show that HeterPS significantly outperforms state-of-the-art approaches in terms of throughput (14.5 times higher) and monetary cost (312.3% smaller).

Concept of heterogeneity index for urban mixed traffic

• This paper conceptualizes a new measure to quantify the heterogeneity of mixed traffic. A proportion-based variant of COV of dynamic PCU values is proposed for this purpose. • This study develops a ‘Gaussian Process Regression’-based classified speed model which is further used to estimate the PCU of different vehicle categories and determine the ‘Heterogeneity Index’ of the mixed traffic stream. • The study proposes the heterogeneity levels using the K-medoids clustering technique along with Davies-Bouldin Index and Silhouette Index as validation measures. • Sensitivity analysis performed in the study shows the effect of traffic volume, traffic composition, and classified speed on the proposed ‘Heterogeneity Index’. Heterogeneity is one of those characteristics which differentiate traffic conditions of a developing country from other developed nations. The heterogeneity which represents the diversity among vehicle categories is suspected to have adverse influences on lane discipline, congestion potential, and road users’ safety. However, the influence of heterogeneity on the above-mentioned parameters has been only measured indirectly by considering traffic composition as an indicator of the prevailed heterogeneity-level. No direct relationship between the heterogeneity and other parameters has been established as there is no methodology available yet for quantifying the heterogeneity of mixed traffic. The present study addresses this problem and conceptualizes the ‘Heterogeneity Index’ (HI) to quantify the heterogeneity present in a mixed traffic stream. HI is conceived as a measure of the dispersion of Passenger Car Units (PCU) for different vehicle categories from its central value. A higher value of HI signifies more diverse vehicle categories present in the traffic stream. PCU of a vehicle category was estimated using the speed-based method and the individual speeds were predicted based on classified volumes using the Gaussian Process Regression model developed in this study. This paper also recommends several categorical levels for easy perception about the intensity of heterogeneity. Further, the sensitivity analysis explored the dynamic aspects of HI. Results showcased how the HI of a traffic stream may vary subject to the combined or the individual change in traffic volume, traffic composition and classified speeds. The outcomes of the study will be useful to estimate the intensity of heterogeneity that prevailed within a mixed traffic stream with varying traffic conditions.

The role of natural resources, clean energy and technology in mitigating carbon emissions in top populated countries

This study highlights the significance of natural resources, clean energy production, and technological advancement to control carbon emissions and to attain COP-26 targets in populous counties of the World. The empirical investigation is based on 30 years (1990–2019) of data analysis of 14 world’s highly populated countries. This study tries to investigate the aforementioned relationship by employing CS-ARDL econometric techniques that mitigate Cross-sectional Dependence and heterogeneity in slop parameters across panels. For this purpose, the study employed some additional independent variables such as economic growth. The empirical findings reveal the facts that high reliance on natural resources to earn more rent worsen the environmental quality. While, clean energy, and technological innovation brings economies toward a clean natural environment through carbon mitigation. The study suggests that the abundance of green natural resources, efficient use of natural resources may help to protect the natural environment even in those countries which are experiencing high population pressure. In addition, the study spot lights that clean energy production and improvement in technology are also helping to achieve the carbon neutrality targets of COP-26 in highly populated economies of the world.

A Decision Support Framework for Pollution Source Detection via Coupled Forward‐Inverse Optimization and Multi‐Information Fusion

AbstractDue to the uncertainty in sensor data, low model accuracy, and high parameter heterogeneity in water quality modeling, pollution source detection (PSD) typically results in a problem of multiple possible solutions, which is the so‐called non‐uniqueness effect. Identifying unique solution to PSD problems is fundamentally essential for water quality control in surface water and groundwater systems. This study proposes a decision support framework to reduce the impact of uncertainty and identify a unique solution using a consensus‐based multiple information fusion method. Multiple water quality information sources are fused in the framework via spatial clustering and temporal Bayesian updating. Considering the real‐world complexity, the recession‐curve displacement method is used to handle multi‐sources scenario to enhance the accuracy of the framework. Meanwhile, a coupled forward‐inverse model instead of random sampling is used to improve the solving efficiency of PSD. The framework is validated by a real water pollution event and a number of semi‐hypothetical case studies. The results show that the prediction accuracy of the single‐source and double‐source pollution problems are 88.40% and 82.69%, with an average relative error of 9.75% and 11.42%, respectively. In addition, the uncertainty contribution quantified by a variance decomposition approach suggests that the interaction between parameter and measurement uncertainties has the greatest impact on the PSD results. The results reveal the advantages of the proposed decision framework for government organizations and communities involved in water environment management.

The impacts of climatic conditions on cereal production: implications for food security in Africa.

Climate change is a confounding factor that affectsfood security in several ways.Although the analyses of earlier studies in this area were largely non-technical, new analytical techniques have been developed tocomprehensively evaluate climate change patternsand their implications for food security. In this study, we use recent developments in panel econometrics, which consider cross-sectional dependence and parameter heterogeneity, to examine the effects of climatic conditions on cereal farming in Africa from 1970Q1 to 2017Q4. The results show that rainfall positively affects cereal crops, although average temperatures are typically unfavourable. In the country-specific scenarios, we observed significant variations in the influence of climatic conditions on cereal production. The causality test results show a two-way causal relationship between climatic conditions-rainfall and temperature-and cereal production. It is suggested that African governments and non-governmental organisations support farmers' adaptation to climate change by implementing policies that prioritise farmers' capacity building and ensure that extension service officers engage with farmers intensively.