Greedy regression in sparse coding space for single-image super-resolution

Investigating single image super-resolution algorithm with deep learning using convolutional neural network for chest digital tomosynthesis

The World Health Organization on March 11, 2020, declared Coronavirus Disease 2019 (Covid-19) a pandemic. Covid-19 is a disease caused by a new type of coronavirus, namely Sars-CoV-2, which affects the respiratory system. Until now, positive confirmed cases of Covid-19 in Indonesia are still occurring every day. This study aims to predict the addition of Covid-19 cases in Indonesia. The data is sourced from the public API page covid19.go.id in the form of an additional number of Covid-19 cases in Indonesia by 122 lines of data. Predictions are made using linear regression and polynomial regression methods as comparisons. Evaluation of the linear regression method obtains a value of R2 = 0.57, while the polynomial regression method obtains a value of R2 = 0.84. Based on these evaluations, the polynomial regression method yields better results than the linear regression method. The prediction of Covid-19 cases in Indonesia from January to March 2022 using the polynomial regression methods predicts that the addition of Covid-19 cases will rise again.

Canopy water interception is a key parameter to study the hydrological cycle, water utilization efficiency, and energy balance in terrestrial ecosystems. Especially in sprinkler-irrigated farmlands, the canopy interception further influences field energy distribution and microclimate, then plant transpiration and photosynthesis, and finally crop yield and water productivity. To reduce the field damage and increase measurement accuracy under traditional canopy water interception measurement, UAVs equipped with multispectral cameras were used to extract in situ crop canopy information. Based on the correlation coefficient (r), vegetative indices that are sensitive to canopy interception were screened out and then used to develop canopy interception models using linear regression (LR), random forest (RF), and back propagation neural network (BPNN) methods, and lastly these models were evaluated by root mean square error (RMSE) and mean relative error (MRE). Results show the canopy water interception is first closely related to relative normalized difference vegetation index (R△NDVI) with r of 0.76. The first seven indices with r from high to low are R△NDVI, reflectance values of the blue band (Blue), reflectance values of the near-infrared band (Nir), three-band gradient difference vegetation index (TGDVI), difference vegetation index (DVI), normalized difference red edge index (NDRE), and soil-adjusted vegetation index (SAVI) were chosen to develop canopy interception models. All the developed linear regression models based on three indices (R△NDVI, Blue, and NDRE), the RF model, and the BPNN model performed well in canopy water interception estimation (r: 0.53–0.76, RMSE: 0.18–0.27 mm, MRE: 21–27%) when the interception is less than 1.4 mm. The three methods underestimate the canopy interception by 18–32% when interception is higher than 1.4 mm, which could be due to the saturation of NDVI when leaf area index is higher than 4.0. Because linear regression is easy to perform, then the linear regression method with NDVI is recommended for canopy interception estimation of sprinkler-irrigated winter wheat. The proposed linear regression method and the R△NDVI index can further be used to estimate the canopy water interception of other plants as well as forest canopy.

Publication bias is a serious problem in meta-analysis. Various methods have been developed to detect the presence of publication bias in meta-analysis. These methods have been assessed and compared in many dichotomous studies utilizing the log-odds ratio as the measure of effect. This paper evaluates and compares the performance of three popular methods, namely the Egger’s linear regression method, the Begg and Mazumdar’s rank correlation method and the Duvall and Tweedie’s trim and fill method, on meta-analysis of continuous data. The data comprised simulated meta-analyses with different levels of primary studies in the absence and presence of induced publication bias. The performance of these methods were measured through the power and type 1 error rate for the tests. The results suggest the trim and fill method to be superior in terms of its ability to detect publication bias when it exists, even in presence of only 5% unpublished studies. However this method is not recommended for large meta-analysis as it produces high rate of false-positive results. Both linear regression and rank correlation method performed relatively well in moderate bias but should be avoided in small meta-analysis as their power is very low in this data.

BackgroundCross-validation techniques in genetic evaluations encounter limitations due to the unobservable nature of breeding values and the challenge of validating estimated breeding values (EBVs) against pre-corrected phenotypes, challenges which the Linear Regression (LR) method addresses as an alternative. Furthermore, beef cattle genetic evaluation programs confront challenges with connectedness among herds and pedigree errors. The objective of this work was to evaluate the LR method's performance under pedigree errors and weak connectedness typical in beef cattle genetic evaluations, through simulation.MethodsWe simulated a beef cattle population resembling the Argentinean Brangus, including a quantitative trait selected over six pseudo-generations with a heritability of 0.4. This study considered various scenarios, including: 25% and 40% pedigree errors (PE-25 and PE-40), weak and strong connectedness among herds (WCO and SCO, respectively), and a benchmark scenario (BEN) with complete pedigree and optimal herd connections.ResultsOver six pseudo-generations of selection, genetic gain was simulated to be under- and over-estimated in PE-40 and WCO, respectively, contrary to the BEN scenario which was unbiased. In genetic evaluations with PE-25 and PE-40, true biases of − 0.13 and − 0.18 genetic standard deviations were simulated, respectively. In the BEN scenario, the LR method accurately estimated bias, however, in PE-25 and PE-40 scenarios, it overestimated biases by 0.17 and 0.25 genetic standard deviations, respectively. In herds facing WCO, significant true bias due to confounding environmental and genetic effects was simulated, and the corresponding LR statistic failed to accurately estimate the magnitude and direction of this bias. On average, true dispersion values were close to one for BEN, PE-40, SCO and WCO, showing no significant inflation or deflation, and the values were accurately estimated by LR. However, PE-25 exhibited inflation of EBVs and was slightly underestimated by LR. Accuracies and reliabilities showed good agreement between true and LR estimated values for the scenarios evaluated.ConclusionsThe LR method demonstrated limitations in identifying biases induced by incomplete pedigrees, including scenarios with as much as 40% pedigree errors, or lack of connectedness, but it was effective in assessing dispersion, and population accuracies and reliabilities even in the challenging scenarios addressed.

Special Section on Enabling Technologies for Network-Based Applications

This paper proposes a single image super resolution algorithm with the aim of satisfying three desirable characteristics, namely, high quality of the produced images, adaptability to image contents...

본 연구에서는 KRX 유가증권시장 전체의 개별 투자자, 기관 투자자, 외국인 투자자들의 거래 활동이 시장 변동성에 미치는 영향을 분석하였다. 특히, 투자자들의 매수 주문불균형과 매도 주문불균형이 시장 변동성에 비대칭적으로 영향을 미치는 지를 분석한다. 이러한 관계를 실증 분석하기 위하여 선형회귀모형과 분위수 회귀모형을 사용하였다. 실증 분석 결과는 다음과 같다. 분위수 회귀모형과 선형회귀모형으로 분석한 결과, 시장 전체에서 개인투자자의 매수(매도) 주문불균형이 클수록 시장 변동성은 증가한다. 둘째, 분위수가 높을 때에만 시장 전체에서 기관투자자의 매수(매도) 주문불균형과 시장 변동성은 유의한 양(+)의 관계를 가진다. 셋째, 유가증권시장 전체에서 외국인투자자의 매도 주문불균형이 클수록 시장 변동성은 증가한다. 하지만 외국인투자자의 매수 주문불균형은 시장 변동성과 유의한 관계를 가지지 않는다. 마지막으로, 분위수 회귀모형과 선형회귀모형으로 분석한 결과를 비교하면, 분위수가 낮을(높을) 때 주문불균형의 계수값은 선형회귀모형의 계수값 보다 작기에(크기에) 선형회귀모형으로 추정 시 주문불균형이 시장변동성에 미치는 영향을 과대(과소) 추정한다는 것을 알 수 있다. 본 연구는 분위수 회귀모형으로 유가증권시장 전체의 투자자 유형별 매수와 매도 주문불균형이 시장 변동성에 비대칭적인 영향을 미치는 지를 분석하였다는 데에 있어 학술적 의의를 가진다.This study examines the impact of trading activities of investor types on KRX stock market and then focuses on examining whether buy and sell side imbalances have an asymmetric effect on market volatility. This study employs quantile regression method and linear regression method to investigate asymmetry phenomenon of stock market volatility. Major findings are as follows: First, this study finds individual investors’ buy imbalance and sell imbalance increase stock market volatility and this paper cannot find asymmetry phenomenon. Second, Institutional investors’ trading imbalance cannot increase stock market volatility using linear regression method. However, this paper finds institutional investors’ buy and sell imbalance increases stock market volatility only at the upper quantiles. Third, only foreigners’ sell side imbalance can increase stock market volatility. Finally, linear regression method underestimates (overestimates) the effects of trading imbalance on KRX stock market volatility at the upper (lower) quantiles.

The time score for 50% cell viability (50% effective time; ET50) is used as the index of skin irritation evaluated by a three-dimensional human skin model, such as TESTSKIN and Vitrolife-Skin. ET50 is conventionally estimated by linear interpolation of measurements at two time points, which yields cell viabilities above and below 50%. This simple method is problematic in that biased estimates are occasionally obtained and confidence intervals cannot be appropriately constructed. We compared four estimation methods including a logistic regression method, a log-time regression method, a linear regression method and a newly proposed two-stage method through a Monte-Carlo simulation study in small sample sizes due to the experimental restrictions. The logistic regression method provides almost unbiased estimates, although the confidence interval for ET50 is occasionally not obtained. The log-time regression method and the linear regression method provide positive biased estimates, although the confidence interval for ET50 is obtained in any case. The two-stage method is reasonable, in which the log-time regression method is adopted only if the logistic regression method cannot construct a confidence interval for ET50.

In this paper, we propose a novel single image super-resolution algorithm that integrates a model-based approach with self-learning deep networks. The proposed method can be adapted to low-resolution (LR) images obtained with real acquisition devices where the point spread function is Gaussian-like. By modelling natural image lines as piece-wise smooth functions and approximating the blurring kernel with B-splines, an intermediate high-resolution (HR) image can be first obtained based on Finite Rate of Innovation theory. A self-supervised deep recursive residual network is then applied to further enhance the reconstruction quality. From the simulation results, our algorithm outperforms other self-learning algorithms and achieves state-of-the-art performance.

Optimization of isotherm models for pesticide sorption on biopolymer-nanoclay composite by error analysis

Omi river flows through Omi-Adio, Iddo Local Government Area (LGA) of Oyo State, Nigeria. The watershed hydrology of the river, over the years, has changed considerably due to the ever increasing anthropogenic activities. The resultant effect is the inundation of farmlands and flooding of residential areas in Omi-Adio township, whenever the river overflowed after heavy rain falls. Unfortunately, as with most Nigerian rivers, no discharge record exists for Omi river, making planning for provision of hydraulic structures difficult. One hundred and three years of available rainfall data and existing five years of discharge record for Ogunpa river at Molete grammar school were used to simulate surface runoff for Omi river, using average proportionality and linear regression methods. The approximation of data was based on established literature, which states that Omi and Ogunpa Rivers have similar geological and morphological characteristics. The simulation was validated with field measurements. The constant of proportionality between the two rivers was found to be 2.467. The constants: 'a' and 'b' in the linear regression equation for Omi River were computed to be 0.306445 and 0.0003036, respectively, with a correlation coefficient of 0.98876. The linear regression method was found to be more appropriate for Omi River at Omi Adio, Ibadan, when compared with field values. The study will assist in providing reasonable data in the design of hydraulic structures such as bridges, concrete channels, culverts and flood control structures for Omi-Adio LGA.

PT XYZ is a company operating in the food industry. The company produces various kinds of food, such as biscuits, wafers, vermicelli and noodles. Noodle products are the most popular products, and there are often significant changes in each period. This causes fluctuations and results in uncontrolled inventory. PT XYZ can avoid this problem if it has a way to estimate noodle demand. Forecasting is a planning method for estimating products in the future. The moving average and linear regression methods are the methods used in this research. Based on the results of the calculations carried out, the method chosen is the linear regression method because the smallest Mean Squared Error (MSE) value is 620370900 and the tracking signal results from the linear regression method are close to zero. The results of forecasting the need for noodle products for the next period were 169,524 pcs. It is hoped that the results of this forecasting can help PT XYZ in carrying out production planning.

The success of modern single image super-resolution (SISR) algorithms is inspired by the development of deep convolutional neural networks (CNNs). However, these CNN-based methods require considerable computation and complexity, making it impossible for these methods to perform real-time calculations in edge devices. Thus, lightweight model design has become a development trend in the super-resolution field, including pruning, quantization, and other methods. The 1-bit quantization is an extreme lightweight method which can reduce the calculation amount of the model in an extreme manner and is friendly to hardware such as edge devices. Most existing binary quantization approaches lead to a large information loss during forward propagation, especially in detailed color information (e.g., edge, texture, and contrast). The loss of color information makes modern binary methods unsuitable for SISR tasks. We think the loss occurs because these methods typically utilize a uniform threshold to quantize the weights and activations. Thus, in this article, we thoroughly analyze the difference between normal classification tasks and SISR tasks, and present a binarization scheme based on local means. The proposed method can maintain more detailed information in feature maps using dynamic thresholds during quantization. Specifically, each value in the full precision activations has a corresponding threshold during the quantization process, and those thresholds are determined by the full precision values of the surroundings. In addition, a gradient approximator is introduced to adaptively optimize the gradient for updating binary weights. We then verify the effectiveness of our method for training binary networks on several SISR benchmarks including VDSR and SRResNet. Experimental results show that the proposed method can outperform the state-of-the-art algorithms to obtain binary networks for image super-resolution with better peak signal-to-noise ratio (PSNR) values and visual quality.

Main idea of this study was to increase efficiency of the EASI ECG method introduced by Dover in 1988 using various regression techniques. EASI was proven to have high correlation with standard 12 lead ECG. Multilayer Perceptron (Artificial Neural Network), Sequential Minimal Optimization Regression and Linear Regression methods were used to improve the quality of the 12-lead electrocardiogram derived from four (EASI) electrodes. Computation of Root Mean Squared Error and Correlation Coefficient was performed to measure the overall result of a given method. The lowest RMSE of 20,90 and the highest correlation coefficient of 0,9858 were obtained using ANN method. Second best result was obtained for SMO Regression Method with Radial Basis Function kernel (RMSE equal to 25,97 and correlation coefficient of 0,9813). The least complex regression method of Linear Regression produced results on a level of 28,35 for RMSE and 0,9741 for correlation coefficient. Results obtained for classic Dover algorithm of deriving 12-lead ECG from EASI electrodes were much worse than those obtained for all regression methods. RMSE of 80,10, which is by around 59,19 higher than RMSE of ANN and correlation coefficient of 0,96, which is by 0,03 lower then correlation coefficient of ANN.KeywordsEASI ECGMultilayer PerceptronSMOregression