Strict Assumptions Research Articles

Achieving Correlated Equilibrium by Studying Opponent’s Behavior Through Policy-Based Deep Reinforcement Learning

Game theory is a very profound study on distributed decision-making behavior and has been extensively developed by many scholars. However, many existing works rely on certain strict assumptions such as knowing the opponent's private behaviors, which might not be practical. In this work, we focused on two Nobel winning concepts, the Nash equilibrium, and the correlated equilibrium. We proposed a policy-based deep reinforcement learning model which instead of just learning the regions for corresponding strategies and actions, it learns why and how the rational opponent plays. With our proposed policy-based deep reinforcement learning model, we successfully reached the correlated equilibrium which maximizes the utility for each player. Depending on the scenario, the equilibrium can reach outside of the Nash equilibrium convex hull to achieve higher utility for the players, while the traditional non-regret algorithms cannot. In addition, we also proposed a mathematical model to inverse the calculation of the correlated equilibrium probability to estimate the rational opponent player's payoff. Through simulations, with limited interaction among the players, we showed that our proposed method can achieve the optimal correlated equilibrium where each player gains an equal or higher utility than the Nash equilibrium.

Non-Negative Transfer Learning With Consistent Inter-Domain Distribution

In this letter, we propose a novel transfer learning approach, which simultaneously exploits the intra-domain differentiation and inter-domain correlation to comprehensively solve the drawbacks many existing transfer learning methods suffer from, i.e., they either are unable to handle the negative samples or have strict assumptions on the distribution. Specifically, the sample selection strategy is introduced to handle negative samples by using the local geometry structure and the label information of source samples. Furthermore, the pseudo target label is imposed to slack the assumption on the inter-domain distribution for considering the inter-domain correlation. Then, an efficient alternating iterative algorithm is proposed to solve the formulated optimization problem with multiple constraints. The extensive experiments conducted on eleven real-world datasets show the superiority of our method over state-of-the-art approaches, i.e., our method achieves 11.23% improvement on the MNIST dataset.

An Intelligent Deployment Policy for Deception Resources Based on Reinforcement Learning

Traditional deception-based cyber defenses (DCD) often adopt the static deployment policy that places the deception resources in some fixed positions in the target network. Unfortunately, the effectiveness of these deception resources has been greatly restricted by the static deployment policy, which also causes the deployed deception resources to be easily identified and bypassed by attackers. Moreover, the existing studies on dynamic deployment policy, which make many strict assumptions and constraints, are too idealistic to be practical. To overcome this limitation, an intelligent deployment policy used to dynamically adjust the locations of these deception resources according to the network security state is developed. Starting with formulating the problem of deception resources deployment, we then model the attacker-defender scenario and the attacker’s strategy. Next, the preliminary screening method that can derive the effective deployment locations of deception resources based on threat penetration graph (TPG) is proposed. Afterward, we construct the model for finding the optimal policy to deploy the deception resources using reinforcement learning and design the Q-Learning training algorithm with model-free. Finally, we use the real-world network environment for our experiments and conduct in-depth comparisons with state-of-the-art methods. Our evaluations on a large number of attacks show that our method has a high defense success probability of nearly 80%, which is more efficient than existing schemes.

Robust Motion Compensation for Event Cameras With Smooth Constraint

Event cameras capture brightness changes at an ultra-high speed and output a series of asynchronous events. In order to form image frames without motion blur from event stream, motion compensation needs to be applied. Unlike other algorithms which either estimate ego-motion or estimate optical flow with strict assumptions, we directly align the curved event trajectories with time-varying motion parameters. By maximizing the image energy with smooth constraint on motion parameters, our algorithm can perform accurately and robustly. We conduct extensive experiments on several public datasets and the recently released event camera, i.e. CeleX-5, the experimental results verify the superiority of the proposed algorithm.

Learning and earning in Nigeria: who works for less

This study quantitatively analyses the differences in levels or size of financial returns to education, by gender in the Nigerian labour market with particular emphasis on estimating gender earning differentials across the country, at zonal levels and at different levels of educational attainments. The analysis is static, employing Mincer's wage equation on a cross sectional dataset. An attempt is made at solving the endogeneity problem that often arise from the violation of the strict exogeneity assumption associated with education, using instrumental variable. The study finds skewed earnings against the female gender, especially in Northern Nigeria. Private returns at all levels of education are lower in the Northern region. Based on the findings, the study suggests that developmental programs aimed at curbing gender and regional inequalities in returns especially for the Northern part of Nigeria be encouraged.

Learning and earning in Nigeria: who works for less

This study quantitatively analyses the differences in levels or size of financial returns to education, by gender in the Nigerian labour market with particular emphasis on estimating gender earning differentials across the country, at zonal levels and at different levels of educational attainments. The analysis is static, employing Mincer's wage equation on a cross sectional dataset. An attempt is made at solving the endogeneity problem that often arise from the violation of the strict exogeneity assumption associated with education, using instrumental variable. The study finds skewed earnings against the female gender, especially in Northern Nigeria. Private returns at all levels of education are lower in the Northern region. Based on the findings, the study suggests that developmental programs aimed at curbing gender and regional inequalities in returns especially for the Northern part of Nigeria be encouraged.

Estimation of Hourly Ground-Level PM2.5 Concentration Based on Himawari-8 Apparent Reflectance

Satellite aerosol optical depth (AOD) is a quantitative parameter frequently used to estimate ground-level fine particulate matters (PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> )at regional to global scales. In this article, Himawari-8 apparent reflectance (top-of-atmosphere reflectance) data were used to estimate the hourly ground-level PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> concentrations (Ref-PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> ) using deep neural networks (DNNs), and comparison was conducted with the AOD-based PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> estimation method (AOD-PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> ). In high-density site areas, the Ref-PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> method was closer to the actual situation and more capable of PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> estimation compared with the AOD-PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> method. The PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> samples used in the AOD-PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> method were less than one-half of the Ref-PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> method due to unavailability of AOD observations, which might be due to strict surface assumptions, cloud detection, and error in the aerosol scheme used in the AOD inversion method. This led to many missing values of AOD-derived PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> in the spatial distribution map of a single day. Moreover, similar hourly variations in PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> were observed for both the methods, and the highest concentration of PM <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</sub> appeared at the junction of Jiangsu, Anhui, and Shandong at 08:00, 09:00, and 10:00 in local time, which gradually decreased at 11:00 and reached to a minimum value at 16:00 and 18:00.

From Large to Vast: Composite Forecasting of Vast-Dimensional Realized Covariance Matrices Using Factor State-Space Models

We propose a dynamic factor state-space model for the prediction of high-dimensional realized covariance matrices of asset returns. Using a block LDL decomposition of the joint covariance matrix of assets and factors, we express the realized covariance matrix of the individual assets similar to an approximate factor model. We model the individual parts, i.e., the factor and residual covariances as well as the factor loadings, independently via a tractable state-space approach. This results in closed-form Matrix-F predictive densities for the distinct covariance elements and Student's t predictive densities for the factor loadings. In an out-of-sample forecasting and portfolio selection exercise we compare the performance of the proposed factor model under different specifications of the residual dynamics. These includes block diagonal residuals based on the GICS sector classifications and strict diagonality assumptions as well as combinations of both using linear shrinkage. We find that the proposed model performs very well in an empirical application to realized covariance matrices for 225 NYSE traded stocks using the well-known Fama-French factors and sector-specific factors represented by Exchange Traded Funds (ETFs).

Mudfold: An R Package for Nonparametric IRT Modelling of Unfolding Processes

Abstract Item response theory (IRT) models for unfolding processes use the responses of individuals to attitudinal tests or questionnaires in order to infer item and person parameters located on a latent continuum. Parametric models in this class use parametric functions to model the response process, which in practice can be restrictive. MUDFOLD (Multiple UniDimensional unFOLDing) can be used to obtain estimates of person and item ranks without imposing strict parametric assumptions on the item response functions (IRFs). This paper describes the implementation of the MUDFOLD method for binary preferential-choice data in the R package mudfold. The latter incorporates estimation, visualization, and simulation methods in order to provide R users with utilities for nonparametric analysis of attitudinal questionnaire data. After a brief introduction in IRT, we provide the method ological framework implemented in the package. A description of the available functions is followed by practical examples and suggestions on how this method can be used even outside the field of psychometrics.

On the Use of Permutation Tests in the Significance Testing of Response Surface Function Parameters

Objective: The methods of experimental design were first used in agricultural experiments performed by R. A. Fisher. The development of experimental design methods took place along with their effective use in production companies. The most frequently used designs of experiments are the factorial designs. One of the stages in the factorial design of experiments is the estimation of the response surface function formula which describes the influence of factors on the response variable values. The aim of this article is to propose a method to indicate the factors which have a significant influence on the response variable. Research Design & Methods: In this case, in the classical approach, the t -test of the significance of particular parameters of the response surface function is used. The t -test requires fulfilment of the assumptions about the distribution and independence of the model errors. If the assumptions are not fulfilled, or the sample size is not sufficient, the use of the t -test is unjustified. An alternative approach to verify the significance of response surface parameters is a permutation test. Permutation tests use simulation methods and do not entail the fulfilment of strict assumptions relating to the distribution of errors and the sample size of experimental data. Findings : The paper deals with the use of a permutation test that allows us to assess the significance of response surface function parameters when the quantity of experimental data is small. These results were obtained using parametric tests and permutation tests. Implications/Recommendations: Based on the performed calculations, it was found that it is possible to use permutation tests to analyse the response surface function, especially when the assumptions about the residuals of the model are not fulfilled or the number of considered experimental trials is small. Contribution: A proper analysis of the response surface function is an important stage in the design of experiments. In the case of a small quantity of experimental data, assessment of the significance of the model and the parameters of the response surface function using parametric tests may lead to incorrect conclusions. Therefore, the use of permutation tests was indicated as an alternative approach in the analysis of the response surface function.

Assessment of survey protocol for estimates of abundance for elusive nocturnal primates

Abstract ContextEstimates of a species abundance and habitat preferences provide vital information on their status and the appropriate conservation management. For nocturnal arboreal primates, obtaining reliable estimates of these parameters is particularly challenging because of their cryptic behaviour, often resulting in a small number of detections. Although techniques are available for assessing the abundance of species with a low probability of detection, most require strict assumptions that are difficult to meet. AimsHere, we aimed to explore the possibility of improving nocturnal-primate abundance estimates when detection probability is low and to determine the minimal effort required to calculate reasonable estimates of their overall abundance and the effect of habitat type on abundance estimates. MethodsWe used count data obtained from spotlighting along line transects for estimating density of Bengal slow loris (Nycticebus bengalensis) in north-eastern Thailand with N-mixture hierarchical modelling, to run simulations of varying survey parameters and asses the effort needed to produce robust estimates based on the relative bias from each simulation. Key resultsN-mixture analysis showed that the data obtained from our study were still biased (9%), with a lambda of 1.79 lorises, detection probability of 0.11, 50 survey sites and a maximum of 12 sample occasions. The simulation results found that increasing the number of sample occasions to 14 per transect would produce an acceptable bias (&amp;lt;5%). ConclusionsWe recommend that future studies on nocturnal arboreal species should use preliminary surveys to gauge the specific lambda and probability of detection so as to establish the effort needed to produce reasonable estimates of abundance. ImplicationsOur study showed that count data obtained from spotlighting can be used to produce robust abundance estimates of nocturnal arboreal species. Unlike simple encounter rate, this method incorporates detection probability and habitat preferences, yet does not require additional trained field technicians.

Event-triggered control for a class of switched uncertain nonlinear systems

This paper investigates the event-triggered tracking control for switched nonlinear systems with parametric uncertainties. Due to the interplay between system switching and event-triggered sampling, two problems arise in the control scheme design. One is the mismatch between subsystems and the corresponding controllers as the switch occurs inside the sampling interval. This problem is rather challenging because the existing methods need to know/assume the maximum mismatch period a priori and use some calculation techniques that are only applicable for linear systems. The other is to avoid the Zeno behavior in the event-triggered control since the switch may trigger additional events while such a switch can be sufficiently close to the last triggering instant. To solve the above problems, we co-design a novel switching event-triggered mechanism and a mode-dependent adaptive control law. With this control scheme, the mismatch behavior is addressed without strict assumptions in the literature, and the absence of Zeno is guaranteed by providing a positive minimum inter-trigger time. The tracking error can be regulated to an adjustable neighborhood of the origin. The obtained results are validated by the tracking control of a single-link robotic manipulator.

Sensitivity-Based Warmstarting for Nonlinear Model Predictive Control With Polyhedral State and Control Constraints

Model predictive control (MPC) is of increasing interest in applications for constrained control of multivariate systems. However, one of the major obstacles to its broader use is the computation time and effort required to solve a possibly nonconvex optimal control problem (OCP) online. This article introduces a sensitivity-based warmstarting strategy for systems with nonlinear dynamics and polyhedral constraints with the goal of reducing the computational footprint of MPC controllers. It predicts changes in the solution of the parameterized OCP as the parameter varies by calculating the semiderivative of the solution mapping. The main novelty of this article is that the polyhedrality of the constraints allows us to avoid imposing any constraint qualification conditions or strict complementarity assumptions. A numerical study featuring MPC applied to unmanned aerial vehicles illustrates the proposed approach.

Semiglobal exponential control of Euler–Lagrange systems using a sliding-mode disturbance observer

Sliding-mode disturbance observer (SMDOB) is appealing for its ability in the estimation of system uncertainties and disturbances without introducing serious chattering in the sliding mode technique. However, existing SMDOB-based controllers with stability guarantee were analyzed by only considering disturbances or under a strict assumption that the time derivatives of system uncertainties are bounded prior to control implementation. This paper proposes a SMDOB-based tracking control strategy for Euler–Lagrange systems with modeling uncertainties and external disturbances. The SMDOB is designed by the construction of an extended state observer embedded by an improved super-twisting algorithm. Compared with existing SMDOB-based controllers, the significant advantage of the proposed controller lies in semiglobal exponential stability guarantee for the combined observer–controller system without the assumption that the time derivatives of modeling uncertainties are bounded by constants. The chattering caused by switching signals is alleviated by passing the switching signals through a low-pass filter in the SMDOB and is compressed by the frequency bandwidths of both the SMDOB and the control system. The effectiveness of the designed SMDOB-based controller is illustrated by simulations on a two-link robot manipulator.

The time buffer approximated Buffer Allocation Problem: A row–column generation approach

One of the main problems in production systems is the buffer sizing. Choosing the right buffer size, at each production stage, that allows to achieve some performance measure (usually throughput or waiting time) is known as Buffer Allocation Problem (BAP), and it has been widely studied in the literature. Due to its complexity, BAP is usually approached using decomposition methods, under very strict system assumptions, or using simulation-optimization techniques. In this paper, the approximated mathematical programming formulation of the BAP simulation-optimization based on the time buffer concept is used. Using this approximation, buffers are modeled as temporal lags (time buffers) and this allows to use Linear Programming (LP) instead of Mixed Integer Linear Programming (MILP) models. Although LP models are easier to solve than MILPs, the huge dimension and the complex solution space topology of the time buffer approximated BAP call for ad hoc solution algorithms. To this purpose, a row-column generation algorithm is proposed, which exploits the theoretical properties of the time buffer approximation to reduce the solution time. The proposed algorithm has been compared with a standard LP solver (ILOG CPLEX) and with a state-of-the-art MILP solver and it proved to be better than the LP solver in most of the cases, and more robust than the MILP solver with respect to computation time. Moreover, the LP model (for flow lines) is able to solve the BAP also for assembly/disassembly lines.

An improved Dai–Kou conjugate gradient algorithm for unconstrained optimization

It is gradually accepted that the loss of orthogonality of the gradients in a conjugate gradient algorithm may decelerate the convergence rate to some extent. The Dai–Kou conjugate gradient algorithm (SIAM J Optim 23(1):296–320, 2013), called CGOPT, has attracted many researchers’ attentions due to its numerical efficiency. In this paper, we present an improved Dai–Kou conjugate gradient algorithm for unconstrained optimization, which only consists of two kinds of iterations. In the improved Dai–Kou conjugate gradient algorithm, we develop a new quasi-Newton method to improve the orthogonality by solving the subproblem in the subspace and design a modified strategy for the choice of the initial stepsize for improving the numerical performance. The global convergence of the improved Dai–Kou conjugate gradient algorithm is established without the strict assumptions in the convergence analysis of other limited memory conjugate gradient methods. Some numerical results suggest that the improved Dai–Kou conjugate gradient algorithm (CGOPT (2.0)) yields a tremendous improvement over the original Dai–Kou CG algorithm (CGOPT (1.0)) and is slightly superior to the latest limited memory conjugate gradient software package CG$$\_ $$DESCENT (6.8) developed by Hager and Zhang (SIAM J Optim 23(4):2150–2168, 2013) for the CUTEr library.

Sufficient dimension folding in regression via distance covariance for matrix‐valued predictors

AbstractIn modern data, when predictors are matrix/array‐valued, building a reasonable model is much more difficult due to the complicate structure. However, dimension folding that reduces the predictor dimensions while keeps its structure is critical in helping to build a useful model. In this paper, we develop a new sufficient dimension folding method using distance covariance for regression in such a case. The method works efficiently without strict assumptions on the predictors. It is model‐free and nonparametric, but neither smoothing techniques nor selection of tuning parameters is needed. Moreover, it works for both univariate and multivariate response cases. In addition, we propose a new method of local search to estimate the structural dimensions. Simulations and real data analysis support the efficiency and effectiveness of the proposed method.

Investigation of the unsteady responses of airfoils due to small-scale motion

Unsteady pressures, forces, and pitching moments generated by foils experiencing vibratory motion in an incompressible, attached flow configuration are studied within this work. Specifically, two-dimensional, unsteady potential flow calculations are performed on Joukowski foils of varying thickness undergoing variable amplitude, small-scale, heaving or pitching motion over a range of reduced frequencies between 0.01 and 100. These calculated results are compared directly to predictions from implementing the Theodorsen model, which treats foils as infinitely thin flat plates that shed a planar sheet of vorticity. The effects of relaxing these seemingly strict assumptions of the Theodorsen model are explored, and focus is placed on results which show deviations from the Theodorsen model. These include altered unsteady responses for finitely thick foils and non-zero mean angle of attack conditions, non-linearity of the wake of shed vorticity, and analysis of the unsteady streamwise force. Further, within the potential flow framework the particular terms which control the unsteady responses are identified.

Evaluating Alternative Correction Methods for Multiple Comparison in Functional Neuroimaging Research.

A significant challenge for fMRI research is statistically controlling for false positives without omitting true effects. Although a number of traditional methods for multiple comparison correction exist, several alternative tools have been developed that do not rely on strict parametric assumptions, but instead implement alternative methods to correct for multiple comparisons. In this study, we evaluated three of these methods, Statistical non-Parametric Mapping (SnPM), 3DClustSim, and Threshold Free Cluster Enhancement (TFCE), by examining which method produced the most consistent outcomes even when spatially-autocorrelated noise was added to the original images. We assessed the false alarm rate and hit rate of each method after noise was applied to the original images.

Unsupervised learning eigenstate phases of matter

Supervised Learning has been successfully used to produce phase diagrams and identify phase boundaries when local order parameters are unavailable. Here, we apply unsupervised learning to this task. By using readily available clustering algorithms, we are able to extract the distinct eigenstate phases of matter within the transverse-field Ising model in the presence of interactions and disorder. We compare our results to those found through supervised learning and observe remarkable agreement. However, as opposed to the supervised procedure, our method requires no strict assumptions concerning the number of phases present, no labeled training data, and no prior knowledge of the phase diagram. We conclude with a discussion of clustering and its limits.