Reduce Estimation Errors Research Articles

Do macroeconomic variables drive exchange rates independently?

The classical practice in exchange rate model estimation is to use bilateral differentials of macroeconomic variables. Empirically, capital may not place equal importance on the economic variables among all countries. Therefore, allowing each country's variable to enter the model independently may reduce estimation error. Based on simulations, we find that higher correlations between the two countries’ variables tend to inflate prediction errors when the estimation uses bilateral differentials. Applying the Sticky Price Monetary Model (SPMM) to a wide range of countries we find supportive evidence for separating the variables, a finding especially relevant for smaller economies.

Hybrid Genetic Algorithm−Based BP Neural Network Models Optimize Estimation Performance of Reference Crop Evapotranspiration in China

Precise estimation of reference evapotranspiration (ET0) is of significant importance in hydrologic processes. In this study, a genetic algorithm (GA) optimized back propagation (BP) neural network model was developed to estimate ET0 using different combinations of meteorological data across various climatic zones and seasons in China. Fourteen climatic locations were selected to represent five major climates. Meteorological datasets in 2018–2020, including maximum, minimum and mean air temperature (Tmax, Tmin, Tmean, °C) and diurnal temperature range (∆T, °C), solar radiation (Ra, MJ m−2 d−1), sunshine duration (S, h), relative humidity (RH, %) and wind speed (U2, m s−1), were first subjected to correlation analysis to determine which variables were suitable as input parameters. Datasets in 2018 and 2019 were utilized for training the models, while datasets in 2020 were for testing. Coefficients of determination (r2) of 0.50 and 0.70 were adopted as threshold values for selection of correlated variables to run the models. Results showed that U2 had the least r2 with ET0, followed by ∆T. Tmax had the greatest r2 with ET0, followed by Tmean, Ra and Tmin. GA significantly improved the performance of BP models across different climatic zones, with the accuracy of GABP models significantly higher than that of BP models. GABP0.5 model (input variables based on r2 > 0.50) had the best ET0 estimation performance for different seasons and significantly reduced estimation errors, especially for autumn and winter seasons whose errors were larger with other BP and GABP models. GABP0.5 model using radiation/temperature data is highly recommended as a promising tool for modelling and predicting ET0 in various climatic locations.

Online state of charge and state of power co-estimation of lithium-ion batteries based on fractional-order calculus and model predictive control theory

Accurate battery modelling is the cornerstone to state of charge (SOC) and state of power (SOP) co-estimation of lithium-ion batteries in electric vehicles. Due to strong battery nonlinearity over a broad frequency range, traditional integer-order models are incapable of capturing complex battery dynamics for SOC and SOP co-estimation. This paper proposes a fractional-order modified moving horizon estimation (FO-mMHE) algorithm and a fractional-order model predictive control (FO-MPC) algorithm. Firstly, a second-order FOM is constructed by performing a series of hybrid pulse tests at different SOC regions, and its model parameters are identified through a particle swarm optimization-genetic algorithm method. Secondly, online SOC estimation is converted into a constrained optimization problem in a past moving horizon and then solved by the FO-mMHE algorithm, which enables fast convergence speed and proactive smoothing of estimation outcomes. Thirdly, the FO-MPC algorithm is devised to manipulate the current sequence in a prediction horizon for maximizing discharge/charge power accumulation and determining battery SOP in real time. Moreover, different battery current–voltage behaviors are comprehensively researched in the prediction horizon over a whole battery operating range. The proposed co-estimation method is validated under different dynamic load profiles. The experimental results demonstrate a SOC estimation error reduction of up to 1.2 % compared with the commonly used fractional-order extended Kalman filter while the SOP estimation error could be limited below 0.35 W.

Advanced modeling method for quantifying cumulative subjective fatigue in mid-air interaction

Interaction in mid-air can be fatiguing. A model-based method to quantify cumulative subjective fatigue for such interaction was recently introduced in HCI research. This model separates muscle units into three states: active (MA) fatigued (MF) or rested (MR) and defines transition rules between states. This method demonstrated promising accuracy in predicting subjective fatigue accumulated in mid-air pointing tasks. In this paper, we introduce an improved model that additionally captures the variations of the maximum arm strength based on arm postures and adds linearly-varying model parameters based on current muscle strength. To validate the applicability and capabilities of the new model, we tested its performance in various mid-air interaction conditions, including mid-air pointing/docking tasks, with shorter and longer rest and task periods, and a long-term evaluation with individual participants. We present results from multiple cross-validations and comparisons against the previous model and identify that our new model predicts fatigue more accurately. Our modeling approach showed a 42.5% reduction in fatigue estimation error when the longitudinal experiment data is used for an individual participant’s fatigue. Finally, we discuss the applicability and capabilities of our new approach.

Spatiotemporal assessments of nutrients and water quality in coastal areas using remote sensing and a spatiotemporal deep learning model

Revealing the spatiotemporal variations of nutrients in coastal waters is crucial to the understanding and evaluation of coastal environment, thereby providing efficient guidance for the aquatic environmental treatment. This study proposed a spatiotemporal-incorporated deep learning model, which is easily applicable to establish the quantitative relationships between measured environmental factors and large-scale satellite maps, and can reduce estimation errors by more than 40% compared with non-spatiotemporal-incorporated deep learning model. The spatiotemporal distributions of dissolved inorganic nitrogen (DIN) and dissolved inorganic phosphate (DIP) over 44400 km2 of the East China Sea on 8-day scale from 2010 to 2018 were obtained. Based on the spatiotemporal variations, the water quality patterns were depicted, and the fluctuation variations of the two essential nutrients were found in the harbors with complex anthropogenic influences, in the typical estuaries with multiple river inputs, and in the open seas with important fisheries. Although the concentration of DIN and DIP decreased by 24% and 19% in 9 years, respectively, the water quality level in the inshore sea has not been significantly improved, especially in autumn and winter. Further, we quantitatively analyzed the main factors of deteriorated water and provided scientific suggestions for targeted monitoring and regional cooperative governances.

Unsupervised Gait Phase Estimation With Domain-Adversarial Neural Network and Adaptive Window.

The performanceof previous machine learning models for gait phase is only satisfactory under limited conditions. First, they produce accurate estimations only when the ground truth of the gait phase (of the target subject) is known. In contrast, when the ground truth of a target subject is not used to train an algorithm, the estimation error noticeably increases. Expensive equipment is required to precisely measure the ground truth of the gait phase. Thus, previous methods have practical shortcoming when they are optimized for individual users. To address this problem, this study introduces an unsupervised domain adaptation technique for estimation without the true gait phase of the target subject. Specifically, a domain-adversarial neural network was modified to perform regression on continuous gait phases. Second, the accuracy of previous models can be degraded by variations in stride time. To address this problem, this study developed an adaptive window method that actively considers changes in stride time. This model considerably reduces estimation errors for walking and running motions. Finally, this study proposed a new method to select the optimal source subject (among several subjects) by defining the similarity between sequential embedding features.

Order Demand Forecast Using a Combined Approach of Stepwise Linear Regression Coefficients and Artificial Neural Network

Abstract
 Nowadays, businesses' forecasts to meet the demands have become more critical. This study aimed to predict the fifteen-day order demand for an order fulfillment center using a Multilayer Perceptron Neural Network (MLPNN). The dataset used in the study was created from a real database of a large Brazilian logistics company and thirteen variables. Linear Regression Coefficients (LRC) were used as a feature selection method to reduce estimation errors. The study showed that among the variables, order type_A (A5), order type_B (A6), and order type_C (A7) had the most significant impact on total order forecasting. The effect of A6 was found to be greater than the effect of A7 and A5. The performance of the proposed model was evaluated using the mean absolute percent error (MAPE). LRC-MLPNN provided a MAPE of 2.97%. The results showed that better forecasting performance was obtained by selecting the independent variables to be used as input to the forecasting model with LRC. The proposed model can also be applied to different estimation problems.

Improving 5G network performance for OFDM-IDMA system resource management optimization using bio-inspired algorithm with RSM

The OFDM-IDMA system is a new era in the wireless domain. It is affected by the MAI and CFO effects. There are no specific CFO reduction techniques that have been identified as the best solution for a multicarrier multiuser system. This paper proposes a bio-inspired scheme for optimizing the performance of an OFDM-IDMA system. To improve BER performance in the presence of CFOs in a multiuser environment, the SIC-MUD and SIC-MUD with SU-LA algorithms are presented. The MAI effect is mitigated by the SIC-MUD technique. The SU-LA algorithm, on the other hand, improves channel estimation performance by optimizing pilot positions. The combination of these algorithms contributes to the reduction of estimation errors and, as a result, ensures the achievement of 0.0472 MSE. The CFO values used in the simulation are 0, 0.1, and 0.2, and the users are 1, 4, 8, with 16 QAM over the Rayleigh channel. When compared to the SIC-MUD algorithm, the proposed algorithm improved the BER by 41.17 percent and can tolerate 0.1 CFO in the presence of 8 users. The analysis’s second-order mathematical regression RSM model has an R2 value of 91.35 percent and accurately predicts the system response. It has been discovered that the proposed SU-LA method outperforms the HS bio-inspired algorithm.

Spatial Scale Effect and Correction of Forest Aboveground Biomass Estimation Using Remote Sensing

Forest biomass is critically important for forest dynamics in the carbon cycle. However, large-scale AGB mapping applications from remote sensing data still carry large uncertainty. In this study, an AGB estimation model was first established with three different remote sensing datasets of GF-2, Sentinel-2 and Landsat-8. Next, the optimal scale estimation result was considered as a reference AGB to obtain the relative true AGB distribution at different scales based on the law of conservation of mass, and the error of the scale effect of AGB estimation at various spatial resolutions was analyzed. Then, the information entropy of land use type was calculated to identify the heterogeneity of pixels. Finally, a scale conversion method for the entropy-weighted index was developed to correct the scale error of the estimated AGB results from coarse-resolution remote sensing images. The results showed that the random forest model had better prediction accuracy for GF-2 (4 m), Sentinel-2 (10 m) and Landsat-8 (30 m) AGB mapping. The determination coefficient between predicted and measured AGB was 0.5711, 0.4819 and 0.4321, respectively. Compared to uncorrected AGB, R2 between scale-corrected results and relative true AGB increased from 0.6226 to 0.6725 for Sentinel-2, and increased from 0.5910 to 0.6704 for Landsat-8. The scale error was effectively corrected. This study can provide a reference for forest AGB estimation and scale error reduction for AGB production upscaling with consideration of the spatial heterogeneity of the forest surface.

Prospects and challenges in the use of models to estimate the influence of crop residue input on soil organic carbon in long-term experiments in Canada

Crop residue input plays a central role in regulating soil organic carbon (SOC) storage. Ten long-term field experiments were used to ascertain the changes in SOC in response to differing rates of crop residues. The amount of C input from crop residues varied significantly between and within sites due to soil-environmental conditions, management and cropping systems. Initial SOC stocks ranged from 45 to 165 t C ha−1 in the 0–30 cm soil depth. Four soil C models, Campbell, ICBM, IPCC Tier 2 steady state (IPCC) and RothC with their default parameterization were used to simulate the SOC. Two model ensemble approaches, a mean (Ens_Avg) and a weighted mean (Ens_Weighted) of the model predictions were also included in the evaluation. Individual model performance was evaluated on the model's ability to capture the SOC stock and change in SOC (ΔSOC). Across sites, the Ens_Weighted approach had the lowest overall RMSE for total SOC (5.2 t C ha−1) and ΔSOC (5.5 t C ha−1) followed by Ens_Avg. Ensemble models which had the lowest bias (PE) and model prediction error (d-index). Within individual models, Campbell and IPCC performed better than other models with the lowest RMSE of 5.8 and 6.3 (t C ha−1) for SOC stocks. In general, ICBM and Campbell underestimated, and RothC overestimated the ΔSOC per unit of C input to the soil with the IPCC model close to the average observations. The use of weighted and average ensemble approaches reduced estimation errors. Nonetheless, our results indicate that regional calibration and validation is warranted for the effective quantification of regional SOC dynamics.

A cross-validation statistical framework for asymmetric data integration.

The proliferation of biobanks and large public clinical data sets enables their integration with a smaller amount of locally gathered data for the purposes of parameter estimation and model prediction. However, public data sets may be subject to context-dependent confounders and the protocols behind their generation are often opaque; naively integrating all external data sets equally can bias estimates and lead to spurious conclusions. Weighted data integration is a potential solution, but current methods still require subjective specifications of weights and can become computationally intractable. Under the assumption that local data are generated from the set of unknown true parameters, we propose a novel weighted integration method based upon using the external data to minimize the local data leave-one-out cross validation (LOOCV) error. We demonstrate how the optimization of LOOCV errors for linear and Cox proportional hazards models can be rewritten as functions of external data set integration weights. Significant reductions in estimation error and prediction error are shown using simulation studies mimicking the heterogeneity of clinical data as well as a real-world example using kidney transplant patients from the Scientific Registry of TransplantRecipients.

Propensity Score Matching Underestimates Real Treatment Effect, in a Simulated Theoretical Multivariate Model

Propensity Score Matching (PSM) is a useful method to reduce the impact of Treatment-Selection Bias in the estimation of causal effects in observational studies. After matching, the PSM significantly reduces the sample under investigation, which may lead to other possible biases (due to overfitting, excess of covariation or a reduced number of observations). In this sense, we wanted to analyze the behavior of this PSM compared with other widely used methods to deal with non-comparable groups, such as the Multivariate Regression Model (MRM). Monte Carlo Simulations are made to construct groups with different effects in order to compare the behavior of PSM and MRM estimating these effects. In addition, the Treatment Selection Bias reduction for the PSM is calculated. With the PSM a reduction in the Treatment Selection Bias is achieved (0.983 [0.982, 0.984]), with a reduction in the Relative Real Treatment Effect Estimation Error (0.216 [0.2, 0.232]), but despite this bias reduction and estimation error reduction, the MRM reduces this estimation error significantly more than the PSM (0.539 [0.522, 0.556], p < 0.001). In addition, the PSM leads to a 30% reduction in the sample. This loss of information derived from the matching process may lead to another not known bias and thus to the inaccuracy of the effect estimation compared with the MRM.

Road grade estimation based on Large-scale fuel consumption data of connected vehicles

Road grade is crucial in vehicle control and emission studies but challenging to obtain in large-scale road networks due to current methods’ expensive deployment costs or limited accuracy. This paper proposed a scale-deployable and cost-efficient road grade estimation solution based on the fuel consumption rate (FCR) difference between flat and graded roads. Real-world road grades from design drawings and 261,814 second-by-second vehicle operating data from 680 light-duty vehicles were collected to examine the proposed method’s performance. Sensitivity tests for vehicle types and sample sizes were conducted. Results show that (1) the proposed method acquired road grade with an accuracy of 0.12% mean absolute error (MAE), (2) in positive vehicle specific power (VSP) bins, a 1% road grade caused an average 16% FCR change, and (3) larger-scale fuel consumption data contributed to reducing estimation error which converged from 0.25% to 0.12% as the segment passes increased from 50 to 400.

Smoothing Linear Multi-Target Tracking Using Integrated Track Splitting Filter

Multi-target tracking (MTT) is a challenging issue due to an unknown number of real targets, motion uncertainties, and coalescence behavior of sensor (such as radar) measurements. The conventional MTT systems deal with intractable computational complexities because they enumerate all feasible joint measurement-to-track association hypotheses and recursively calculate the a posteriori probabilities of each of these joint hypotheses. Therefore, the state-of-art MTT system demands bypassing the entire joint data association procedure. This research work utilizes linear multi-target (LM) tracking to treat feasible target detections followed by neighbored tracks as clutters. The LM integrated track splitting (LMITS) algorithm was developed without a smoothing application that produces substantial estimation errors. Smoothing refines the state estimation in order to reduce estimation errors for an efficient MTT. Therefore, we propose a novel Fixed Interval Smoothing LMITS (FIsLMITS) algorithm in the existing LMITS algorithm framework to improve MTT performance. This algorithm initializes forward and backward tracks employing LMITS separately using measurements collected from the sensor in each scan. The forward track recursion starts after the smoothing. Therefore, each forward track acquires backward multi-tracks that arrived from upcoming scans (future scans) while simultaneously associating them in a forward track for fusion and smoothing. Thus, forward tracks become more reliable for multi-target state estimation in difficult cluttered environments. Monte Carlo simulations are carried out to demonstrate FIsLMITS with improved state estimation accuracy and false track discrimination (FTD) in comparison to the existing MTT algorithms.

Digital H∞ filter design for a low frequency multiple‐input multiple‐output system with multirate measurements

AbstractIn this article, a digital filter is designed for a low frequency multiple‐input multiple‐output (MIMO) system with multirate measurements from multiple samplers. A delay compensation is introduced to reduce estimation errors from the multiple samplers and a zero‐order holder. A single‐rate error system is obtained by converting the low frequency MIMO system using a discrete‐time lifting technique. A digital filter is proposed based on a delay‐partitioning method to guarantee asymptotic stability and performance for the single‐rate error system. A numerical example is provided to show effectiveness and superiority of the digital filter in low frequency domain.

Over-the-Air Statistical Estimation

We study schemes and lower bounds for distributed minimax statistical estimation over a Gaussian multiple-access channel (MAC) under squared error loss. Our framework combines statistical estimation and wireless communication. First, we develop “analog” joint estimation-communication schemes that exploit the superposition property of the Gaussian MAC. We characterize their risk in terms of the number of nodes and dimension of the parameter space. Then, we derive information-theoretic lower bounds on the minimax risk of any estimation scheme that is restricted to communicate the samples over a given number of uses of the channel. This shows that the risk achieved by our proposed schemes is within a logarithmic factor of these lower bounds. We compare both achievability and lower bound results to previous “digital” lower bounds, where nodes transmit errorless bits at the Shannon capacity of the MAC. Our key finding is that analog estimation schemes that leverage the physical layer offer a drastic reduction in estimation error over digital schemes relying on a physical-layer abstraction.

Haphazard Intentional Sampling in Survey and Allocation Studies on COVID-19 Prevalence and Vaccine Efficacy.

Haphazard intentional sampling is a method developed by our research group for two main purposes: (i) sampling design, where the interest is to select small samples that accurately represent the general population regarding a set of covariates of interest; or (ii) experimental design, where the interest is to assemble treatment groups that are similar to each other regarding a set of covariates of interest. Rerandomization is a similar method proposed by K. Morgan and D. Rubin. Both methods intentionally select good samples but, in slightly different ways, also introduce some noise in the selection procedure aiming to obtain a decoupling effect that avoids systematic bias or other confounding effects. This paper compares the performance of the aforementioned methods and the standard randomization method in two benchmark problems concerning SARS-CoV-2 prevalence and vaccine efficacy. Numerical simulation studies show that haphazard intentional sampling can either reduce operating costs in up to to achieve the same estimation errors yielded by the standard randomization method or, the other way around, reduce estimation errors in up to using the same sample sizes.

Shrinking beta

Beta is used in many applications, ranging from asset pricing tests to cost of capital estimation, investment management and risk management. Beta needs to be estimated, and shrinkage to its cross-sectional average value of 1 is often applied to reduce estimation error. Since beta is the product of the return correlation of a security with the market and its return volatility relative to that of the market, we shrink correlation and volatility separately and evaluate the predictive power of this approach. We find economically and statistically significant gains from applying more shrinkage to correlations than to volatilities.

Rauch-Tung-Striebel Smoothing Linear Multi-Target Tracking in Clutter

Most conventional data association based multi-target tracking (MTT) algorithms typically suffer from intractable computational complexities and could not perform in an environment where large number of closely spaced multiple targets move across each other in clutters. Unlike to the existing MTT systems, the linear multi-target (LM) algorithm modifies the measurement detection followed by neighbored tracks as a clutter, hence, it updates the track without the influence of other tracks. Thus, LM technique is a computationally efficient algorithm that allows the multi-target system to play like a single target tracking algorithm. Smoothing maximizes the state estimation accuracy and reduces estimation error based on future scan measurement. However, only few research paper focused on the LM algorithm without utilizing the benefits of the smoothing. This paper presents Rauch-Tung-Striebel Smoothing in the linear multi-target based on integrated probabilistic data association (RTS-LMIPDA). The RTS-LMIPDA algorithm fuses forward and backward LM track predictions to obtain the smoothing prediction which is required to calculate the smoothing multi-target state estimates in the forward track. Numerical analysis is presented to illustrate the estimation accuracy and false track discrimination (FTD) performances of RTS-LMIPDA in comparison to the existing MTT algorithms using the simulations.

Methodology to filter out outliers in high spatial density data to improve maps reliability

The considerable volume of data generated by sensors in the field presents systematic errors; thus, it is extremely important to exclude these errors to ensure mapping quality. The objective of this research was to develop and test a methodology to identify and exclude outliers in high-density spatial data sets, determine whether the developed filter process could help decrease the nugget effect and improve the spatial variability characterization of high sampling data. We created a filter composed of a global, anisotropic, and an anisotropic local analysis of data, which considered the respective neighborhood values. For that purpose, we used the median to classify a given spatial point into the data set as the main statistical parameter and took into account its neighbors within a radius. The filter was tested using raw data sets of corn yield, soil electrical conductivity (ECa), and the sensor vegetation index (SVI) in sugarcane. The results showed an improvement in accuracy of spatial variability within the data sets. The methodology reduced RMSE by 85 %, 97 %, and 79 % in corn yield, soil ECa, and SVI respectively, compared to interpolation errors of raw data sets. The filter excluded the local outliers, which considerably reduced the nugget effects, reducing estimation error of the interpolated data. The methodology proposed in this work had a better performance in removing outlier data when compared to two other methodologies from the literature.