Developing a new chance constrained NDEA model to measure the performance of humanitarian supply chains

Data envelopment analysis (DEA) is a method for measuring performance of decision making units (DMUs). Conventional DEA models view DMUs as black boxes. Network DEA (NDEA) models have been developed to overcome this shortfall. This paper develops a new NDEA model based on modified enhanced Russell measure model. This paper measures performance of humanitarian supply chains (HSCs) by an NDEA model. Capabilities of the proposed model are addressed by theorems. However, in the real world, there might be stochastic data. This paper presents a stochastic version of the proposed NDEA model to measure the performance of HSCs. We analyse main properties of our model. We present a case study to demonstrate the applicability of the proposed model.     

Prediction of the human response time with the similarity and quantity of information

Memory is one of brain processes that are important when trying to understand how people process information. Although a large number of studies have been made on the human performance, little is known about the similarity effect in human performance. The purpose of this paper is to propose and validate the quantitative and predictive model on the human response time in the user interface with the concept of similarity. However, it is not easy to explain the human performance with only similarity or information amount. We are confronted by two difficulties: making the quantitative model on the human response time with the similarity and validating the proposed model by experimental work. We made the quantitative model based on the Hick's law and the law of practice. In addition, we validated the model with various experimental conditions by measuring participants' response time in the environment of computer-based display. Experimental results reveal that the human performance is improved by the user interface's similarity. We think that the proposed model is useful for the user interface design and evaluation phases.     

On a consistent field transfer in non linear inelastic analysis and ultimate load computation

In this work we propose a field transfer operator for remeshing carried out in the course of incremental analysis of a non linear inelastic behavior. The proposed procedure is geared towards the ultimate load computation of a complex structure, where we choose the appropriate mesh grading for each different phase of computations, starting with a coarse mesh for the initial linear response and going towards a more refined mesh for highly nonlinear inelastic response. The proposed projection operator is developed on the basis of diffuse approximation method. The key feature of such an operator is to guarantee the conservation of relevant mechanics quantities which ensures a superior performance of the proposed field transfer with respect to the standard remeshing procedure. We present the illustrative results both for an isotropic damage model and standard plasticity model, indicating very satisfying performance.     

Stacked Attention Networks for Referring Expressions Comprehension

Referring expressions comprehension is the task of locating the image region described by a natural language expression, which refer to the properties of the region or the relationships with other regions. Most previous work handles this problem by selecting the most relevant regions from a set of candidate regions, when there are many candidate regions in the set these methods are inefficient. Inspired by recent success of image captioning by using deep learning methods, in this paper we proposed a framework to understand the referring expressions by multiple steps of reasoning. We present a model for referring expressions comprehension by selecting the most relevant region directly from the image. The core of our model is a recurrent attention network which can be seen as an extension of Memory Network. The proposed model capable of improving the results by multiple computational hops. We evaluate the proposed model on two referring expression datasets: Visual Genome and Flickr30k Entities. The experimental results demonstrate that the proposed model outperform previous state-of-the-art methods both in accuracy and efficiency. We also conduct an ablation experiment to show that the performance of the model is not getting better with the increase of the attention layers.     

Phase pattern denoising using a regularized cost function with complex-valued Markov random fields based on a discrete model

We present a simple and effective method for denoising phase patterns based on a discrete model. The proposed filtering method transforms the image denoising problem to solving the energy diffusion problem of a system with complex-valued fields. We establish an appropriate cost function that uses the discrete form of complex-valued Markov random fields. The attractiveness of the proposed filtering method includes three points: the first is that the filtering process can be easily implemented using an iterative method, the second is that 2π phase jumps are well preserved, and the third is its little computational effort. The performance of the proposed method is demonstrated by simulated and experimentally obtained phase patterns.     

Design of predictable production scheduling model using control theoretic approach

As one of the most important planning and operational issues in manufacturing systems, production scheduling generally deals with allocating a set of resources over time to perform a set of tasks. Recently, control theoretic approaches based on nonlinear dynamics of continuous variables have been proposed to solve production scheduling problems as an alternative to traditional production scheduling methods that deal with decision-making components in discrete nature. The major goal of this paper is to improve predictability and performance of an existing scheduling model that employs the control theoretic approach, called distributed arrival time controller (DATC), to manage arrival times of parts using an integral controller. In this paper, we first review and investigate unique dynamic characteristics of the DATC in regards to convergence and chattering of arrival times. We then propose a new arrival time controller for the DATC that can improve predictability and performance in production scheduling. We call the new mechanism the double integral arrival-time controller (DIAC). We analyse unique characteristics of the DIAC such as oscillatory trajectory of arrival times, their oscillation frequency, and sequence visiting mechanism. In addition, we compare scheduling performance of the DIAC to the existing DATC model through computational experiments. The results show that the proposed system can be used as a mathematical and simulation model for designing adaptable manufacturing systems in the future.     

Developing performance measurement system for Internet of Things and smart factory environment

To cope with large fluctuations in the demand of a commodity, it is necessary for the manufacturing system to have rapid reactive ability. This requirement may be secured by performance measurement. Although manufacturing companies have used information systems to manage performance, there has been the difficulty of capturing real-time data to depict real situations. The recent development and application of the Internet of Things (IoT) has enabled the resolution of this problem. In demonstration of the functionality of IoT, we developed an IoT-based performance model consistent with the ISA-95 and ISO-22400 standards, which define manufacturing processes and performance indicator formulas. The development comprised three steps: (1) Selection of the Key Performance Indicators of the Overall Equipment Effectiveness (OEE), and the development of an IoT-based production performance model, (2) Implementation of the IoT-based architecture and performance measurement process using Business Process Modelling and (3) Validation of the proposed model through virtual factory simulation. We investigated the effect of the IoT-workability on the OEE, based on the final results of the simulation, both for the planned and actual productions. The simulation results showed that the proposed model represented the timestamp data acquired by IoT and captured the entire production process, thus enabling the determination of real-time performance indicators.     

The Plant Location and Technology Acquisition Problem

This paper presents an analytical approach for simultaneous optimization of the plant location, capacity acquisition and technology selection decisions in a multi-product environment. The proposed approach can be useful when there is considerable interaction between these structural decisions e.g., in global manufacturing companies. We present a formal definition of the plant location and technology acquisition problem and provide a mathematical model. We describe the analytical properties of the model, which lead to the development of a solution algorithm. Progressive piecewise linear underestimation constitutes the backbone of our solution algorithm. The arising subproblems are amenable to a dual-based approach. We report on a set of experiments that improved our understanding of the interaction among facility design decisions and showed that the computational performance of the proposed solution procedure is quite satisfactory.     

Gaussian Process Single-Index Models as Emulators for Computer Experiments

A single-index model (SIM) provides for parsimonious multidimensional nonlinear regression by combining parametric (linear) projection with univariate nonparametric (nonlinear) regression models. We show that a particular Gaussian process (GP) formulation is simple to work with and ideal as an emulator for some types of computer experiment as it can outperform the canonical separable GP regression model commonly used in this setting. Our contribution focuses on drastically simplifying, reinterpreting, and then generalizing a recently proposed fully Bayesian GP-SIM combination. Favorable performance is illustrated on synthetic data and a real-data computer experiment. Two R packages, both released on CRAN, have been augmented to facilitate inference under our proposed model(s).     

Old dog,new tricks: a modelling view of simple moving averages

Simple moving average (SMA) is a well-known forecasting method. It is easy to understand and interpret and easy to use, but it does not have an appropriate length selection mechanism and does not have an underlying statistical model. In this paper, we show two statistical models underlying SMA and demonstrate that the automatic selection of the optimal length of the model can easily be done using this finding. We then evaluate the proposed model on a real data-set and compare its performance with other popular simple forecasting methods. We find that SMA performs better both in terms of point forecasts and prediction intervals in cases of normal and cumulative values.     

Stochastic cyclic flow lines with blocking: Markovian models

We consider a cyclic flow line model that repetitively produces multiple items in a cyclic order. We examine performance of stochastic cyclic flow line models with finite buffers of which processing times have exponential or phase-type distributions. We develop an exact method for computing a two-station model by making use of the matrix geometric structure of the associated Markov chain. We present a computationally tractable approximate performance computing method that decomposes the line model into a number of two-station submodels and parameterizing the submodels by propagating the starvation and blocking probabilities through the adjacent submodels. We discuss performance characteristics including comparison with random order processing and effects of the job variation and the job processing sequence. We also report the accuracy of our proposed method.     

A lightweight deep learning model for retinal optical coherence tomography image classification

Automated retinal disease detection and grading is one of the most researched areas in medical image analysis. In recent years, Deep Learning models have attracted much attention in this field. Hence, in this paper, we present a Deep Learning-based, lightweight, fully automated end-to-end diagnostic system for the detection of the two major retinal diseases, namely diabetic macular oedema (DME) and drusen macular degeneration (DMD). Early detection of these diseases is important to prevent vision impairment. Optical coherence tomography (OCT) is the main imaging technique for detecting these diseases. The model proposed in this work is based on residual blocks and channel attention modules. The performance of the model is evaluated using the publicly available Mendeley OCT dataset and the Duke dataset. We were able to achieve a classification accuracy of 99.5% in the Mendeley test dataset and 94.9% in the Duke dataset with the proposed model. For the application, we performed an extensive evaluation of pre-trained models (LeNet, AlexNet, VGG-16, ResNet50 and SE-ResNet). The proposed model has a much smaller number of trainable parameters and shows superior performance compared to existing methods.     

A Method for Performance Degradation Modeling Based upon the Accelerated Experiment

Performance degradation modeling plays an important role in prognostics and health management of mechanical system. Influenced by the complex structure of the hydraulic pump and the limited experiment standards, it is hard to establish an appropriate performance degradation model. To fulfill current requirements, a method for establishing the performance degradation model based on accelerated experiment is proposed. In order to describe the general trend of the degradation, the double-stress exponential model is firstly established as the theoretical degradation model. On this basement, combined with the characteristics of the experiment, the accelerating coefficient is settled; meanwhile, the procedures for assuring the model parameters are presented. Furthermore, based on the accelerated experiment of the hydraulic pump under various stresses, the performance degradation model is finally established. Result of the experimental analysis indicates that the proposed method is applicable and the presented model is effective to measure the performance degradation of pump.     

Ejector performance analysis under overall operating conditions considering adjustable nozzle structure

The variable-geometry ejector (VGE) is feasible for unstable heat-source utilization; the ejector can be adjusted to its design point to obtain high efficiency. Moreover, as the adjustable nozzle in the VGE significantly affects the performance of the ejector, a theoretical model is necessary to evaluate VGE performance. In this study, a two-dimensional theoretical model was proposed based on an adjustable-nozzle theory. Method of characteristics was employed to accurately predict the driving-flow development in the mixing section. In addition, the suction-flow velocity distribution on the effective area was considered. The proposed model was validated by employing the data from literature and additional experimental data obtained from a VGE test setup using R134a. The validation result shows that the proposed model predicts the ejector performance accurately; moreover, the model is more adaptive while the nozzle configuration changes. The theoretical model, proposed herein, is practical for the design and application of the VGE.     

A condition-based maintenance policy for stochastically deteriorating systems

We focus on the analytical modeling of a condition-based inspection/replacement policy for a stochastically and continuously deteriorating single-unit system. We consider both the replacement threshold and the inspection schedule as decision variables for this maintenance problem and we propose to implement the maintenance policy using a multi-level control-limit rule.In order to assess the performance of the proposed maintenance policy and to minimize the long run expected maintenance cost per unit time, a mathematical model for the maintained system cost is derived, supported by the existence of a stationary law for the maintained system state.Numerical experiments illustrate the performance of the proposed policy and confirm that the maintenance cost rate on an infinite horizon can be minimized by a joint optimization of the maintenance structure thresholds, or equivalently by a joint optimization of a system replacement threshold and the aperiodic inspection schedule.     

Likelihood-based inference for a frailty-copula model based on competing risks failure time data

A competing risks phenomenon arises in industrial life tests, where multiple types of failure determine the working duration of a unit. To model dependence among marginal failure times, copula models and frailty models have been developed for competing risks failure time data. In this paper, we propose a frailty-copula model, which is a hybrid model including both a frailty term (for heterogeneity among units) and a copula function (for dependence between failure times). We focus on models that are useful to investigate the reliability of marginal failure times that are Weibull distributed. Furthermore, we develop likelihood-based inference methods based on competing risks data, including accelerated failure time models. We also develop a model-diagnostic procedure to assess the adequacy of the proposed model to a given dataset. Simulations are conducted to demonstrate the operational performance of the proposed methods, and a real dataset is analyzed for illustration. We make an R package “gammaGumbel” such that users can apply the suggested statistical methods to their data.     

Supply chain relationship quality and performance in technological turbulence: an artificial neural network approach

A well-functioning supply chain management relationship cannot only develop seamless coordination with valuable members, but also improve operational efficiency to secure greater market share, increased profits and reduced costs. An accurate decision-making system considering multifactor relationship quality is highly desired. This study offers an alternative perspective and characterisation of the supply chain relationship quality and performance. A decision-making model is proposed with an artificial neural network approach for supply chain continuous performance improvement. Supply chain performance is analysed via a supervised learning back-propagation neural network. An ‘inverse’ neural network model is proposed to predict the supply chain relationship quality conditions. Optimal performance parameters can be obtained using the proposed neural network scheme, providing significant advantages in terms of improved relationship quality. This study demonstrates a new solution with the combination of qualitative and quantitative methods for performance improvement. The overall accuracy rate of the decision-making model is 88.703%. The results indicated that trust has the greatest influence on the supply chain performance. Relationship quality among supply chain partners impacts performance positively as the pace of technological turbulence increases.     

Performance of a laser Earth-to-satellite link over turbulence and beam wander using the modulated gamma-gamma irradiance distribution

We present an analytical framework for the performance evaluation of laser satellite uplinks over the major probabilistic impairments, i.e., atmospheric turbulence and beam wander. Specifically, we consider a ground-station-to-space laser uplink with a Gaussian beam wave model, and we focus on the particular regime assuming untracked beams where beam wandering takes place. In that regime, the modulated gamma-gamma distribution has been proposed as an effective irradiance model to characterize the combined effect of turbulence and beam wander. First we provide a closed-form expression of the probability density function and deduce the fundamental statistics of the new model. Then we evaluate the performance of the laser system assuming coherent detection for several modulation schemes. An appropriate set of numerical results is presented to verify the accuracy of the derived expressions.     

Research on Sensor Network Coverage Enhancement Based on Non-Cooperative Games

Coverage is an important issue for resources rational allocation, cognitive tasks completion in sensor networks. The mobility, communicability and learning ability of smart sensors have received much attention in the past decade. Based on the deep study of game theory, a mobile sensor non-cooperative game model is established for the sensor network deployment and a local information-based topology control (LITC) algorithm for coverage enhancement is proposed. We both consider revenue of the monitoring events and neighboring sensors to avoid nodes aggregation when formulating the utility function. We then prove that the non-cooperative game is an exact potential game in which Nash Equilibrium exists. The proposed algorithm focuses on the local information of the neighboring sensors and decides sensors’ next action based on the actions of the other sensors, which maximizes its own utility function. We finally evaluate the performance of the proposed method through simulations. Simulation results demonstrate that the proposed algorithm can enlarge the coverage of the entire monitoring area while achieving effective coverage of the events.     

An IoT Environment Based Framework for Intelligent Intrusion Detection

Software-defined networking (SDN) represents a paradigm shift in network traffic management. It distinguishes between the data and control planes. APIs are then used to communicate between these planes. The controller is central to the management of an SDN network and is subject to security concerns. This research shows how a deep learning algorithm can detect intrusions in SDN-based IoT networks. Overfitting, low accuracy, and efficient feature selection is all discussed. We propose a hybrid machine learning-based approach based on Random Forest and Long Short-Term Memory (LSTM). In this study, a new dataset based specifically on Software Defined Networks is used in SDN. To obtain the best and most relevant features, a feature selection technique is used. Several experiments have revealed that the proposed solution is a superior method for detecting flow-based anomalies. The performance of our proposed model is also measured in terms of accuracy, recall, and precision. F1 rating and detection time Furthermore, a lightweight model for training is proposed, which selects fewer features while maintaining the model’s performance. Experiments show that the adopted methodology outperforms existing models.