Exponential Weighting Research Articles

Spectral decision analysis and evaluation in an experimental environment for cognitive wireless networks

Dynamic spectrum allocation is the crucial aspect in Cognitive Radio Networks (CRN) to improve spectral efficiency where the selection of the target frequency channel plays a relevant role in the communication performance of the secondary user. Therefore, the current research work aims to analyze and comparatively evaluate three multi-criteria techniques in spectrum decision within an experimental environment. In order to achieve the above, it is incorporated actual information from licensed users through realistic spectral occupancy data measured in the GSM bands, the performance of the proposed methodology is validated through three multi-criteria decision-making techniques: Multiplicative Exponential Weighting (MEW), Simple Additive Weighting (SAW), and Multi-Criteria Optimization and Compromise Solution (VIKOR). The results obtained show the outstanding performance of the multicriteria algorithms in decision making and their application in cognitive radio networks, highlighting the high performance of the SAW algorithm in contrast to MEW and VIKOR. The contribution of this work is the use of actual spectral occupancy information captured through a spectral measurement campaign in Bogota, the use of three evaluation metrics in terms of handoff rate, bandwidth, and throughput, as well as a detailed analysis of the results obtained.

A new measurement association mapping strategy for DOA tracking

To solve the problem of the time-varying direction of arrival (DOA) in uniform linear array (ULA), a novel DOA tracking method based on generalized label multi-Bernoulli (GLMB) filter is proposed. Since the measurement value is a superimposed data, which will lead to the mismatch of measurement association mapping (MAP) in the GLMB filter updated step. To solve this issue, we propose a new measurement association mapping (NMAP) strategy, which redefines the MAP between the target and the measurement. Furthermore, particle filter is given to approximate the posterior distribution of the GLMB filter, and the likelihood function is replaced by the multi-signal classification (MUSIC) spatial spectrum function. Finally, by exponential weighting of the likelihood function, the number of particles in the high likelihood region increases, making the pruning and merging of the GLMB filter more effective. Compared with the existing methods, the proposed method is superior to other algorithms in tracking the source's state and estimating the number of sources.

Ultra-short-term Wind Power Output Prediction Based On LS-WMC Combined Model

In this paper, an ultra short term wind power output prediction method based on least square and weighted Markov chain model is proposed. Firstly, the wind farm output is fitted based on the least square method, and then the prediction accuracy is improved by using the weighted Markov chain model. At the same time, the partial exponential weighting method is used to improve the prediction accuracy of ultra short-term wind power output on the basis of reducing the generation frequency of transition probability matrix, so as to optimize and improve the prediction results. The results show that the method can improve the prediction accuracy.

Multi-Maneuvering Sources DOA Tracking With Improved Interactive Multi-Model Multi-Bernoulli Filter for Acoustic Vector Sensor (AVS) Array

To solve the problem of the direction of arrival (DOA) maneuvering in acoustic vector sensor (AVS) array, we propose an interactive multi-model Multi-target Multi-Bernoulli (IMM-MeMBer) two-dimensional (2-D) DOA tracking algorithm. The idea of the IMM algorithm is employed to interactively estimate the predicted sampled particles at the expense of calculating the likelihood function of the predicted particles by pseudo-spectrum of multi-signal classification (MUSIC). The proposed method fully considers various possibilities of target motion, which makes the target state estimation more effective. Moreover, the de-noising and exponential weighting of MUSIC pseudo-spectrum improve the likelihood function of particles, thus enhancing the weight of particles in high likelihood region and making these particles more effective in the subsequent resampling process. Simulation results show that the IMM-MeMBer algorithm has better performance than projection approximation subspace tracking of deflation MUSIC (PASTD-MUSIC), MeMBer, particle filter (PF), and random finite set PF (RFS-PF) in the DOA tracking of maneuvering sources.

Fully Automated 3-D Ultrasound Segmentation of the Placenta, Amniotic Fluid, and Fetus for Early Pregnancy Assessment.

Volumetric placental measurement using 3-D ultrasound has proven clinical utility in predicting adverse pregnancy outcomes. However, this metric cannot currently be employed as part of a screening test due to a lack of robust and real-time segmentation tools. We present a multiclass (MC) convolutional neural network (CNN) developed to segment the placenta, amniotic fluid, and fetus. The ground-truth data set consisted of 2093 labeled placental volumes augmented by 300 volumes with placenta, amniotic fluid, and fetus annotated. A two-pathway, hybrid (HB) model using transfer learning, a modified loss function, and exponential average weighting was developed and demonstrated the best performance for placental segmentation (PS), achieving a Dice similarity coefficient (DSC) of 0.84- and 0.38-mm average Hausdorff distances (HDAV). The use of a dual-pathway architecture improved the PS by 0.03 DSC and reduced HDAV by 0.27 mm compared with a naïve MC model. The incorporation of exponential weighting produced a further small improvement in DSC by 0.01 and a reduction of HDAV by 0.44 mm. Per volume inference using the FCNN took 7-8 s. This method should enable clinically relevant morphometric measurements (such as volume and total surface area) to be automatically generated for the placenta, amniotic fluid, and fetus. The ready availability of such metrics makes a population-based screening test for adverse pregnancy outcomes possible.

Energy aware tree‐based sink relocation routing protocol to improve the lifetime of wireless sensor networks

SummaryIn wireless sensor network (WSN), the design of routing protocols focuses mainly on the energy efficiency based on the application of the network. Over the past years, researchers developed many solutions to protract the life expectancy of the network by reducing the consumption of energy by each node. One of the main approaches is the usage of sink in mobility; this reduces the hotspot problem and energy consumption. Aiming at this problem, the paper proposes a new routing algorithm which supports the sink mobility with energy aware tree‐based sink relocation routing protocol (TBSR) using adaptive fuzzy decision‐making approach (AF‐MADM). In TBSR, a self ‐organized routing tree is formed, which reduces the consumption of energy. In the beginning, the clustering process is employed using fuzzy possibilistic C‐means (FPCM) algorithm. Further, the fuzzy analytic hierarchy process (FAHP) with multiplicative exponential weighting model is developed for optimal decision making to select the root node. Then the selected root node creates a routing tree based on the distance and energy of the sensor nodes. Then each node uses its cluster information and the distance of the neighbors to select the parent node. For relocating the sink node, the proposed protocol calculates the energy along with hop count and path capacity. These methods make TBSR an energetic and analogous protocol. Results show that the proposed TBSR improves on the energy consumption and the lifetime of the network compared with existing protocols. The simulation is carried out in OMNET++ platform.

Multiple fading factors-based strong tracking variational Bayesian adaptive Kalman filter

If the system model or the statistical characteristics of noise are inaccurate, the past measurements will directly affect the accuracy of current state estimation or even lead to filtering divergence. To overcome above difficulties, a multiple fading factors-based strong tracking variational Bayesian adaptive Kalman filter is proposed. Firstly, the inverse Wishart distribution is adopted to model the measurement noise covariance matrix. Secondly, the remodified measurement noise covariance matrix and the innovation covariance matrix estimated by exponential weighting method are employed to construct the scalar fading factor. Next, the multiple fading factors are calculated to correct the predicted error covariance matrix. Finally, the local optimal estimations of measurement noise covariance matrix and state are obtained by variational Bayesian approach. The target tracking simulations verify that the proposed algorithm has better tracking ability for the predicted error covariance matrix and the measurement noise covariance matrix compared with the existing filters.

A Deviation-Based Detection Method Against False Data Injection Attacks in Smart Grid

State estimation plays a vital role to ensure safe and reliable operations in smart grid. Intelligent attackers can carefully design a destructive and stealthy false data injection attack (FDIA) sequence such that commonly used weighted least squares estimator combined with residual-based detection method is vulnerable to the FDIA. To effectively defend against an FDIA, in this paper, we propose a robust deviation-based detection method, in which an additional Kalman filter is introduced while retaining the original weighted least squares estimator, so that there are two state estimators. Moreover, an exponential weighting function is also applied to the introduced Kalman filter in our proposed method. When an FDIA occurs, the estimation results of weighted least squares estimator depend only on meter measurements at each time slot, but there is an adjustment process of estimated states for the Kalman filter based on historical states' transitions. Meanwhile, based on the exponential weighting function, estimated measurements in the Kalman filter can be adaptively suppressed for different attack strengths of FDIAs, and then the difference of the results of these two estimators increases. Subsequently, FDIAs can be effectively detected by checking the deviation of estimated measurements about the two estimators with a detection threshold. Experimental results validate the effectiveness of the proposed detection method against FDIAs. The impact of different attack strengths and noise on detection performance is also evaluated and analyzed.

A Hybrid Precipitation Index Inspired by the SPI, PDSI, and MCDI. Part I: Development of the Index

AbstractThe filtering properties of the standardized precipitation index (SPI), the Palmer drought severity index (PDSI), and the model calibrated drought index (MCDI) are investigated to determine their relations to past, present, and future precipitation anomalies in regions with a wide diversity of precipitation characteristics. All three indices can be closely approximated by weighted averages of precipitation, but with different weighting. The SPI is well represented by one-sided, uniformly weighted averages; the MCDI is well represented by one-sided, exponentially weighted averages; and the PDSI is well represented by two-sided, exponentially weighted averages with much higher weighting of past and present precipitation than future precipitation. Detailed analyses identify interpretational complications and other undesirable features in the SPI and PDSI. In addition, the PDSI and MCDI are each restricted to single regionally specific “intrinsic” time scales that can significantly differ between the two indices. Inspired by the strengths of the SPI, PDSI, and MCDI, a hybrid index is developed that consists of exponentially weighted averages of past and present precipitation that are implicit in the PDSI and MCDI. The explicit specification of the exponential weighting allows users to control the time scale of the hybrid index to investigate precipitation variability on any time scale of interest. This advantage over the PDSI and MCDI is analogous to the controllability of the time scale of the SPI, but the exponentially fading memory is more physical than the uniform weighting of past and present precipitation in the SPI.

Latency estimation for chronic disease risk: a damped exponential weighting model

Identifying the susceptible period when environmental factors affect disease risk is essential for understanding disease etiology. Most existing epidemiologic studies use oversimplified summaries of time-dependent exposures such as baseline or most current exposure, or the cumulative average of exposure over available follow-up periods. In this paper, we introduce a damped exponential weighting model for estimating optimal exposure weights for different time intervals. This model can accommodate flexible patterns of weights and can be fit using standard software. We applied the model to assess the latency of BMI and alcohol for post-menopausal breast cancer based on 30-year exposure history in the Nurses' Health Study. We have also performed a simulation study to assess the validity of the proposed hypothesis testing and estimation procedures in realistic conditions. We found that the type I error is close to 0.05; the bias in our parameter estimates is low and the coverage probability of interval estimates is close to 0.95. For ER+/PR+ breast cancer we found that recent BMI was a more important predictor of risk than more distant BMI; for ER-/PR- breast cancer, no latency was found and risk was characterized by cumulative high levels of BMI over a long period of time. For alcohol intake, we saw a strong positive association with cumulative intake for ER+/PR+ breast cancer; no significant association was found for cumulative intake or for any latency measure of risk for ER-/PR- breast cancer. Our results underscore the value of an easy-to-implement approach to latency analysis of exposure profiles for chronic disease.

Location Preferences of New Pedestrian Bridges Based on Multi-Criteria Decision-Making and GIS-Based Estimation

Non-motorized pedestrian and bicycle traffic is an effective and efficient tool for reducing the negative environmental impacts of transport and improving the quality of life in urban conditions. The strategies of creating new attractive spaces on the waterfront are prevalent amongst the municipalities in different countries. This kind of development intends the construction of new connection bridges, usually meant solely for walking and cycling. There are a significant number of studies covering the theme of pedestrian bridges, but the studies typically focus on technical parameters – serviceability, stress and vibrations, specifications for the design. Researchers and stakeholders rarely discuss displacement strategy, expenditure and future usability. This study aims to find out the best and the most useful bridge locations that would contribute to pedestrian network improvement, would add value to city image and give other benefits. A novel hybrid Multi-Criteria Decision-Making (MCDM) model, based on five different multi-criteria decision-making methods: Multiplicative Exponential Weighting (MEW), method of Evaluation Based on Distance from Average Solution (EDAS), an Additive Ratio Assessment (ARAS) method, expert judgement, and Step‐Wise Weight Assessment Ratio Analysis (SWARA), is presented. A developed model allows solving complicated problems and finding a rationally, balanced solution. Arguments derived from this study help politicians and town planners as well as society.

Exponential weights in multivariate regression and a low-rankness favoring prior

Nous etablissons des garanties theoriques pour l’erreur de prediction de l’agregat par poids exponentiels dans le cadre de la regression multivariee, c’est-a-dire lorsqu’on souhaite predire une etiquette multidimensionnelle. Nous considerons le modele de regression a design fixe et a bruit borne. La premiere consequence de nos garanties est qu’il n’est pas necessaire d’exiger l’independance des observations : une condition de symetrie sur la seule distribution du bruit suffit pour obtenir des bornes precises. Ce resultat necessite que les vecteurs de regression soient bornees. Une seconde consequence curieuse concerne le cas des vecteurs de regression non bornes mais independant de bruit. Il s’avere que l’application des poids exponentiels aux vecteurs d’etiquette perturbes par un bruit uniforme conduit a un estimateur satisfaisant une inegalite d’oracle exacte. La derniere contribution est l’instanciation des inegalites d’oracle proposees a des problemes dans lesquels le parametre inconnu est une matrice. Nous proposons une loi a priori favorisant les matrices a faible rang et montrons qu’elle conduit a un estimateur optimal sous des hypotheses faibles.

Iterative Identification of Hammerstein Parameter Varying Systems With Parameter Uncertainties Based on the Variational Bayesian Approach

The identification of the multiple model-based Hammerstein parameter varying systems is studied in this paper. The parameters of the considered systems vary as the systems perform on different operating conditions. For each local model, the input nonlinear output-error structure is introduced to describe the dynamical property. Allocating an exponential weighting function to each local model, the nonlinear dynamics of the global system is approximated by combining all local models. The variational Bayesian (VB) approach is adopted to find the solution to the problem of parameter estimation. For the parameter uncertainties, instead of the point estimation, the posterior distribution of each model parameters is obtained under the framework of the VB approach. Two numerical simulation examples and an experiment carried on a multitank system have been employed to demonstrate that the proposed approach can work effectively.

A variational Bayesian approach for robust identification of linear parameter varying systems using mixture laplace distributions

The robust identification problem of the linear parameter varying (LPV) systems with output data corrupted by outliers is considered in this paper. The local identification approach is used, and the LPV model is obtained by interpolating the local models with an exponential weighting function. In order to handle outliers that could occur in industrial processes, the corresponding probabilistic model is established with the process noise assumed to be mixture Laplace distributed, then the formulas to iteratively update the unknown model parameters and noise-free output are derived under the Variational Bayesian (VB) framework, which approximates the required posteriors and could avoid high dimensional integrals. One numerical example and a practical chemical process are employed to verify the efficacy of the developed algorithm.

Assessing the indicators of economic growth in Indonesia using spatial principal components analysis

The objective of this study is to reduce the number of indicators that affect correlated economic growth into the main component variables. The variables used Gross Regional Domestic Product (GRDP) (X1), Foreign Direct Investment (X2), Domestic Direct Investment (X3), Employment (X4), Human Community (X5), Export (X6), and Import (X7). The data used are secondary data taken from Statistics Indonesia in 2018. This study uses the Geographical Weighted Principal Component Analysis (GWPCA) with the kernel weighting used is exponential kernel weighting. The results of this study are the main component variables or new variables that can be explained about 89% of the original variables. Based on the first main component (PC1) which consists of three groups of variables that affect economic growth indicators, namely: the area of which economic growth indicators are influenced by the Human Resources variable (X5) followed by the GRDP variable (X1), the region whose indicators of economic growth are influenced by the Export variable (X6) is followed by the Import variable (X7), and the region which is an indicator of economic growth is influenced by the Import variable (X7) is accepted by the Export variable (X6).

DOA Tracking Based on Unscented Transform Multi-Bernoulli Filter in Impulse Noise Environment.

Aiming at the problem of multiple-source direction of arrival (DOA) tracking in impulse noise, this paper models the impulse noise by using the symmetric α stable (SαS) distribution, and proposes a DOA tracking algorithm based on the Unscented Transform Multi-target Multi-Bernoulli (UT-MeMBer) filter framework. In order to overcome the problem of particle decay in particle filtering, UT is adopted to select a group of sigma points with different weights to make them close to the posterior probability density of the state. Since the α stable distribution does not have finite covariance, the Fractional Lower Order Moment (FLOM) matrix of the received array data is employed to replace the covariance matrix to formulate a MUSIC spatial spectra in the MeMBer filter. Further exponential weighting is used to enhance the weight of particles at high likelihood area and obtain a better resampling. Compared with the PASTD algorithm and the MeMBer DOA filter algorithm, the simulation results show that the proposed algorithm can more effectively solve the issue that the DOA and number of target are time-varying. In addition, we present the Sequential Monte Carlo (SMC) implementation of the UT-MeMBer algorithm.

An Application of Manifold Learning in Global Shape Descriptors

With the rapid expansion of applied 3D computational vision, shape descriptors have become increasingly important for a wide variety of applications and objects from molecules to planets. Appropriate shape descriptors are critical for accurate (and efficient) shape retrieval and 3D model classification. Several spectral-based shape descriptors have been introduced by solving various physical equations over a 3D surface model. In this paper, for the first time, we incorporate a specific manifold learning technique, introduced in statistics and machine learning, to develop a global, spectral-based shape descriptor in the computer graphics domain. The proposed descriptor utilizes the Laplacian Eigenmap technique in which the Laplacian eigenvalue problem is discretized using an exponential weighting scheme. As a result, our descriptor eliminates the limitations tied to the existing spectral descriptors, namely dependency on triangular mesh representation and high intra-class quality of 3D models. We also present a straightforward normalization method to obtain a scale-invariant and noise-resistant descriptor. The extensive experiments performed in this study using two standard 3D shape benchmarks—high-resolution TOSCA and McGill datasets—demonstrate that the present contribution provides a highly discriminative and robust shape descriptor under the presence of a high level of noise, random scale variations, and low sampling rate, in addition to the known isometric-invariance property of the Laplace–Beltrami operator. The proposed method significantly outperforms state-of-the-art spectral descriptors in shape retrieval and classification. The proposed descriptor is limited to closed manifolds due to its inherited inability to accurately handle manifolds with boundaries.

A brief survey on the choice of parameters for: “Kernel density estimation for time series data”

This paper presents an overview of the empirical performance of some of the common methods for parameter selection in the area of enhanced dynamic kernel density and distribution estimation with exponentially declining weights. It is shown that exponential weighting delivers accurate nonparametric density and quantile evaluations, without common corrections for scale and/or location in most of the financial time series considered, provided that parameters are chosen appropriately with computationally heavy Least-Squares routines. For more time-efficient numerical optimisations and/or simple kernel adaptive estimation strategies, Least-Squares routines may be re-written with exponentially weighted binned kernel estimators. This insures equally effective parameters evaluation under the different choices of kernel functional forms, though binning strategy becomes an important component of estimations. On the other hand, it is also highlighted that if the estimations target is to mine time-varying nonparametric quantiles, kernel functional forms and bandwidths may not be necessary for these evaluations. Combining exponential weights with empirical distribution estimator provides a very similar quantile performance to the kernel enhanced estimator, while parametric specifications may provide a better extreme quantiles outlook.

Study on reflectivity data interpolation and mosaics for multiple Doppler weather radars

Multiple weather radars’ joint detection can enlarge the radar detection range and can improve the detection accuracy, which has become an important method to monitor large-scale weather effectively. In this paper, the methods of interpolating reflectivity volume scan data onto the grids in the Cartesian coordinate system and three-dimensional mosaic for gridded reflectivity data of several radars are studied. In order to retain the physical characteristics of the raw data, the smoothing parameters of adaptive Barnes interpolation are improved which based on the space structure of the raw reflectivity data. Through comparison of constant altitude plan position indication (CAPPI) images obtained by the commonly used interpolation schemes and the improved Adaptive Barnes interpolation, it was found that the latter can provide consecutive reflectivity fields and retain high-resolution structure comparable to the raw data. The mean deviation filter filtering erroneous grid data and the data are fused by various fusion methods. Result shows that Exponential Weighting is an excellent method which can provide continuous three-dimensional reflectivity mosaic data.

Weighted preliminary-summation-based principal component analysis for non-Gaussian processes

For industrial chemical process, preliminary-summation-based principal component analysis (PS-PCA), an amended PCA method was recently provided for coping with both Gaussian and non-Gaussian characteristics. By summing the training and monitoring data respectively, PS-PCA is capable of resolving the issue of non-Gaussian processes and achieves higher fault detection rate than the traditional PCA. However, in the PS-PCA summation operation, all data samples are regarded as the same weight, which results in the fault information of newly-samples may be diluted, leading to significant detection delays. To address this challenge, in this paper, we propose a novel weighted PS-PCA (WPS-PCA) method that employs an exponential weighting scheme to put more emphasis on recent information. Subsequently, a mathematical argument demonstrates that when the number of variables is enough plentiful, the obtained summation combined with the generalized central limit theorem conforms to approximately a Gaussian distribution. The kurtosis relationships indicate this conversion will bring out well-pleasing feasibility for conventional PCA. Ultimately, the proposed technique verifies detection performance using the Tennessee Eastman process, which is compared with the existing PCA and PS-PCA schemes, in terms of the fault detection time and fault detection rate. The simulation studies reveal that the proposed method is efficient and superior.