Comparison of ARIMA and LSTM for prediction of hemorrhagic fever at different time scales in China

Objectives This study intends to build and compare two kinds of forecasting models at different time scales for hemorrhagic fever incidence in China. Methods Autoregressive Integrated Moving Average (ARIMA) and Long Short-Term Memory Neural Network (LSTM) were adopted to fit monthly, weekly and daily incidence of hemorrhagic fever in China from 2013 to 2018. The two models, combined and uncombined with rolling forecasts, were used to predict the incidence in 2019 to examine their stability and applicability. Results ARIMA (2, 1, 1) (0, 1, 1)12, ARIMA (1, 1, 3) (1, 1, 1)52 and ARIMA (5, 0, 1) were selected as the best fitting ARIMA model for monthly, weekly and daily incidence series, respectively. The LSTM model with 64 neurons and Stochastic Gradient Descent (SGDM) for monthly incidence, 8 neurons and Adaptive Moment Estimation (Adam) for weekly incidence, and 64 neurons and Root Mean Square Prop (RMSprop) for daily incidence were selected as the best fitting LSTM models. The values of root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) of the models combined with rolling forecasts in 2019 were lower than those of the direct forecasting models for both ARIMA and LSTM. It was shown from the forecasting performance in 2019 that ARIMA was better than LSTM for monthly and weekly forecasting while the LSTM was better than ARIMA for daily forecasting in rolling forecasting models. Conclusions Both ARIMA and LSTM could be used to build a prediction model for the incidence of hemorrhagic fever. Different models might be more suitable for the incidence prediction at different time scales. The findings can provide a good reference for future selection of prediction models and establishments of early warning systems for hemorrhagic fever.

The Fitting Optimization Path Analysis on Scale Missing Data: Based on the 507 Patients of Poststroke Depression Measured by SDS

Objective. To explore the optimal fitting path of missing data of the Scale to make the fitting data close to the real situation of patients’ data. Methods. Based on the complete data set of the SDS of 507 patients with stroke, the data simulation sets of Missing Completely at Random (MCAR), Missing at Random (MAR), and Missing Not at Random (MNAR) were constructed by R software, respectively, with missing rates of 5%, 10%, 15%, 20%, 25%, 30%, 35%, and 40% under three missing mechanisms. Mean substitution (MS), random forest regression (RFR), and predictive mean matching (PMM) were used to fit the data. Root mean square error (RMSE), the width of 95% confidence intervals (95% CI), and Spearman correlation coefficient (SCC) were used to evaluate the fitting effect and determine the optimal fitting path. Results. when dealing with the problem of missing data in scales, the optimal fitting path is ① under the MCAR deletion mechanism, when the deletion proportion is less than 20%, the MS method is the most convenient; when the missing ratio is greater than 20%, RFR algorithm is the best fitting method. ② Under the Mar mechanism, when the deletion ratio is less than 35%, the MS method is the most convenient. When the deletion ratio is greater than 35%, RFR has a better correlation. ③ Under the mechanism of MNAR, RFR is the best data fitting method, especially when the missing proportion is greater than 30%. In reality, when the deletion ratio is small, the complete case deletion method is the most commonly used, but the RFR algorithm can greatly expand the application scope of samples and save the cost of clinical research when the deletion ratio is less than 30%. The best way to deal with data missing should be based on the missing mechanism and proportion of actual data, and choose the best method between the statistical analysis ability of the research team, the effectiveness of the method, and the understanding of readers.

Temporal trends in the physical fitness of Hong Kong adolescents between 1998 and 2015

Low physical fitness in adolescence is linked with increased cardiometabolic risk and early all-cause mortality. This study aimed to estimate temporal trends in the physical fitness of Hong Kong adolescents aged 12–17 years between 1998 and 2015. Physical fitness (9-min run/walk, sit-ups, push-ups, and sit-and-reach) and body size data in a total of 28,059 adolescents tested across five population-representative surveys of Hong Kong secondary school pupils, were reported. Temporal trends in means were estimated at the gender-age level by best-fitting sample-weighted linear regression, with national trends estimated by a post-stratified population-weighting procedure. Overall, there were small declines in 9-min run/walk (effect size (ES) = 0.29 (95%CI: 0.32, 0.26)) and sit-ups performance (ES = 0.24 (95%CI: 0.27, 0.21)), with negligible changes in push-ups and sit-and-reach performance. There were small concurrent increases in both mean height and body mass, with a negligible increase in sum of skinfolds. Trends in mean physical fitness and body size/ were not always uniform across the population distribution. The small declines in mean 9-min run/walk and sit-ups performance for Hong Kong adolescents are suggestive of corresponding declines in cardiorespiratory fitness and abdominal/core endurance, respectively. Increased national health promotion strategies are required to improve existing fitness trends.

Analog and Photon Signal Splicing for CO2-DIAL Based on Piecewise Nonlinear Algorithm

In the CO2 differential absorption lidar (DIAL) system, signals are simultaneously collected through analog detection (AD) and photon counting (PC). These two kinds of signals have their own characteristics. Therefore, a combination of AD and PC signals is of great importance to improve the detection capability (detection range and accuracy) of CO2-DIAL. The traditional signal splicing algorithm cannot meet the accuracy requirements of CO2 inversion due to unreasonable data fitting. In this paper, a piecewise least square splicing algorithm is developed to make signal splicing more flexible and efficient. First, the lidar signal is segmented, and according to the characteristics of each signal, the best fitting parameters are obtained by using the least square fitting with different steps. Then, all the segmented and fitted signals are integrated to realize the effective splicing of the near-field AD signal and the far-field PC signal. A weight gradient strategy is also adopted in signal splicing, and the weights of the AD and PC signals in the spliced signal change with the height. The splicing effect of the improved algorithm is evaluated by the measured signal, which are obtained in Wuhan, China, and the splice of the AD and PC signals in the range of 800–1500 m are completed. Compared with the traditional method, the evaluation parameter R2 and the residual sum of squares of the spliced signal are greatly improved. The linear relationship between the AD and PC signals is improved, and the fitting R2 of differential absorption optical depth reaches 0.909, indicating that the improved signal splicing algorithm can well splice the near-field AD signal and the far-field PC signal.

Comparison of Body Mass Index with Modified Subjective Global Assessment for Detection of Malnutrition Among Patients Under Maintenance Hemodialysis: A Multi-Center Cross-Sectional Study

Background: Patients under maintenance hemodialysis are at increased risk of malnutrition, causing from multitude of factors. Present study aims to assess the prevalence of malnutrition among maintenance hemodialysis patients using both modified subjective global assessment score and body mass index, compare them and assess the sensitivity and specificity of body mass index for detecting malnutrition, along with determining a new cutoff value for BMI that better represent the maintenance hemodialysis patient’s nutritional status. Methods: This was a cross-sectional study conducted in the hemodialysis unit of Bangabandhu Sheikh Mujib Medical University, Sir Salimullah Medical College Mitford Hospital, BIRDEM General Hospital and National Institute of Kidney Diseases & Urology; among 80 adult CKD patients who were on regular (≥2 sessions per week) maintenance hemodialysis for more than 3 months without any acute infection, during the period of July 2016 to June 2017. Nutritional assessment was done for each patient using modified SGA score along with BMI. Sensitivity analysis of WHO recommended cutoff value for BMI was done among the study population using modified SGA score as gold standard test for detection of malnutrition among the respondents. ROC curve was used to estimate the best fitting cutoff value of BMI that showed highest sensitivity, specificity and accuracy for detracting malnutrition among maintenance hemodialysis patients. Results: The study participants were predominantly male (66.3%) and from age group 45 to 59 years (36.3%). Modified SGA score detected 90.0% of the study population as malnourished. WHO recommended 18.5 kg/m2 cutoff value was also used to detect malnutrition among study population and 13.8% were found to be malnourished, with a sensitivity and specificity of 12.5% and 75.0% respectively. Accuracy was found to be 18.8%. Using ROC curve, 23.1 kg/m2 was found to be the best fitting cutoff value of BMI for the study population to detect malnutrition. With a sensitivity of 47.2%, specificity of 37.5% and accuracy of 46.3%. Conclusion: BMI showed low sensitivity for detecting malnutrition among patients under maintenance hemodialysis, compared to modified SGA score and should be avoided as a screening tool, but 23.1 kg/m2 cutoff value for BMI showed potential to be used as an easy to use and quick tool for detecting malnutrition among such patients. Further study with larger sample size could shed more light on this. JOPSOM 2021; 40(1): 14-21

Nonlinear models in the height description of the Rhino sunflower cultivar

ABSTRACT: Sunflower produces achenes and oil of good quality, besides serving for production of silage, forage and biodiesel. Growth modeling allows knowing the growth pattern of the crop and optimizing the management. The research characterized the growth of the Rhino sunflower cultivar using the Logistic and Gompertz models and to make considerations regarding management based on critical points. The data used come from three uniformity trials with the Rhino confectionery sunflower cultivar carried out in the experimental area of the Federal University of Santa Maria - Campus Frederico Westphalen in the 2019/2020 agricultural harvest. In the first, second and third trials 14, 12 and 10 weekly height evaluations were performed on 10 plants, respectively. The data were adjusted for the thermal time accumulated. The parameters were estimated by ordinary least square’s method using the Gauss-Newton algorithm. The fitting quality of the models to the data was measured by the adjusted coefficient of determination, Akaike information criterion, Bayesian information criterion, and through intrinsic and parametric nonlinearity. The inflection points (IP), maximum acceleration (MAP), maximum deceleration (MDP) and asymptotic deceleration (ADP) were determined. Statistical analyses were performed with Microsoft Office Excel® and R software. The models satisfactorily described the height growth curve of sunflower, providing parameters with practical interpretations. The Logistics model has the best fitting quality, being the most suitable for characterizing the growth curve. The estimated critical points provide important information for crop management. Weeds must be controlled until the MAP. Covered fertilizer applications must be carried out between the MAP and IP range. ADP is an indicator of maturity, after reaching this point, the plants can be harvested for the production of silage without loss of volume and quality.

Development of a New Shoe Fitting for Obese Adults

Background: Standard fit as well as wide-fit footwear not currently being pertinent and comfortable for the obese adults. The biometric measurements of obese foot (such as foot length, foot width, heel girth, instep girth, waist girth, and ball girth,) significantly differ from healthy adults. Aim: This study aims to develop a new shoe fitting for obese adults based on significant relationships among the relevant biometric parameters of the foot. Method: These measurements of obese foot were determined using a Brannock device and measuring tape. All kinds of foot girth measurements were analyzed against scaling based on foot width or current fitting, BMI, foot length, heel girth, instep girth, waist girth and ball girth and compared these data with ANOVA. Result: Results showed that responses of all kinds of girths against waist girth scaling provide best fitting prospects of obese adults than current standard fit as well as other parameters. Conclusion: From the study and results, it can be concluded that shoe fitting based on waist girth can give more precise comfort and improve the ergonomic fitness of the product for obese users. Bangladesh Journal of Medical Science Vol. 21(1) 2022 Page : 45-53

Adsorptive Removal of Arsenic by Synthetic Iron-loaded Goethite: Isotherms, Kinetics, and Mechanism

Arsenic contamination in the groundwater is a worldwide concern. Therefore, this study was designed to use synthetic iron-loaded goethite to remove arsenic. Adsorption was significantly pH-dependent; hence, pH values between 5.0 and 7.0 resulted in the highest removal of arsenate and arsenite. Langmuir and Freundlich isotherms were almost perfectly matched in terms of strong positive coefficient of determination “R2” arsenate – 0.941 and 0.992 and arsenite – 0.945 and 0.993. The adsorption intensity “n” resulted as arsenate – 2.542 and arsenite – 2.707; besides separation factor “RL” found as arsenate – 0.1 and arsenite – 0.5, respectively. However, both “n” and “RL” leads to a favourable adsorption process. Temkin isotherm yielded in equal binding energies “bt” showing as 0.004 (J/μg) for both arsenate and arsenite. Jovanovic monolayers isotherm was dominated by the Langmuir isotherm. This resulting in maximum adsorption capacity “Qmax” of arsenate – 1369.877 and arsenite – 1276.742 (μg/g), which approaches to the saturated binding sites. Kinetic data revealed that adsorption equilibrium was achieved in 240 – arsenate and 360 – arsenite (minutes), respectively. Chemisorption was found effective with high “R2” values 0.981 ­– arsenate and 0.994 – arsenite, respectively, with the best fitting of pseudo-second order. Moreover, Brunauer Emmett Teller (BET), Scanning Electron Microscopy (SEM), X-ray diffraction (XRD), and Fourier Transform Infrared Spectroscopy (FTIR) were used to determine the morphological content, surface area, crystalline structure, and chemical characteristics of the adsorbent. It is anticipated that optimal arsenic removal was achieved by the porosity, chemical bindings, and surface binding sites of the adsorbent.

The network of the subjective experience in embodiment phenomena

Body illusions are designed to temporarily alter body representation by embodying fake bodies or part of them. Despite their large use, the embodiment questionnaires have been validated only for the embodiment of fake hands in the Rubber Hand Illusion (RHI). With the current study, we aimed at (i) extending the validation of embodiment questionnaires to a different illusory situation (e.g., the Full-Body Illusion - FBI); (ii) comparing two methods to explore the questionnaires structures: a classic Exploratory Factor Analysis (EFA) and a modern Exploratory Graph Analysis (EGA). 118 healthy participants completed an FBI procedure where the subjective experience of embodiment was measured with a standard questionnaire. The EFA results in two-factor structures. However, the Confirmatory Factor Analysis (CFA) fit indices do not show a good fit with the data. Conversely, the EGA identified four communities: ownership, agency, co-location and disembodiment; the solution was confirmed by a CFA. Overall, the EGA seems to be the best fitting method for the present data. Our results confirm the EGA as a suitable substitute for a more classical EFA. Moreover, the emerged structure suggests that the FBI induces similar effects to the RHI, implying that the embodiment sensations are common to different illusory methods. Public Significance Statement: The study indicates that the Exploratory Graph Analysis is a convenient substitute for a more classical Exploratory Factor Analysis. Moreover, the present paper suggests that different illusory methods induce similar embodiment effects.

Application of reinforcement learning for the optimization of clinch joint characteristics

Due to increasing challenges in the area of lightweight design, the demand for time- and cost-effective joining technologies is steadily rising. For this, cold-forming processes provide a fast and environmentally friendly alternative to common joining methods, such as welding. However, to ensure a sufficient applicability in combination with a high reliability of the joint connection, not only the selection of a best-fitting process, but also the suitable dimensioning of the individual joint is crucial. Therefore, few studies already investigated the systematic analysis of clinched joints usually focusing on the optimization of particular tool geometries against shear and tensile loading. This mainly involved the application of a meta-model assisted genetic algorithm to define a solution space including Pareto optima with all efficient allocations. However, if the investigation of new process configurations (e. g. changing materials) is necessary, the earlier generated meta-models often reach their limits which can lead to a significantly loss of estimation quality. Thus, it is mainly required to repeat the time-consuming and resource-intensive data sampling process in combination with the following identification of best-fitting meta-modeling algorithms. As a solution to this problem, the combination of Deep and Reinforcement Learning provides high potentials for the determination of optimal solutions without taking labeled input data into consideration. Therefore, the training of an Agent aims not only to predict quality-relevant joint characteristics, but also at learning a policy of how to obtain them. As a result, the parameters of the deep neural networks are adapted to represent the effects of varying tool configurations on the target variables. This provides the definition of a novel approach to analyze and optimize clinch joint characteristics for certain use-case scenarios.