Residual Mean Square Error Research Articles

The historical lepto-variance of the US stock returns

<abstract> <p>Regression trees (RT) involve sorting samples based on a particular feature and identifying the splitting point that yields the highest drop in variance from a parent node to its children. The optimal factor for reducing mean squared error (MSE) is the target variable itself. Consequently, employing the target variable as the basis for splitting sets an upper limit on the reduction of MSE and, equivalently, a lower limit on the residual MSE. Building upon this observation, we define lepto-regression as the process of constructing an RT of a target feature on itself. Lepto-variance pertains to the portion of variance that cannot be mitigated by any regression tree, providing a measure of inherent variance at a specific tree depth. This concept is valuable as it offers insights into the intrinsic structure of the dataset by establishing an upper boundary on the "resolving power" of RTs for a sample. The maximal variance that can be accounted for by RTs with depths up to <italic>k</italic> is termed the sample k-bit macro-variance. At each depth, the overall variance within a dataset is thus broken into lepto- and macro-variance. We perform 1- and 2-bit lepto-variance analysis for the entire US stock universe for a large historical period since 1926. We find that the optimal 1-bit split is a 30–70 balance. The two children subsets are centered roughly at −1% and 0.5%. The 1-bit macro-variance is almost 42% of the total US stock variability. The other 58% is structure beyond the resolving power of a 1-bit RT. The 2-bit lepto-variance equals 26.3% of the total, with 42% and 47% of the 1-bit lepto-variance of the left and right subtree, respectively.</p> </abstract>

Enhanced adsorption of toluene using nano silver-modified activated carbon derived from Ricinodendron heudolitii shells: A comparative study"

Considering the risk of toluene exposure in water resources, researchers have been prompted to conduct experiments to remove them from aqueous solutions. This study investigates the use of Ricinodendron Heudolitii shells as a precursor for extracting an adsorbent (activated carbon) through chemical activation with phosphoric acid (H3PO4). The obtained activated carbon was then modified with Quercetagetin stabilized silver nanoparticles (Qtg-Ag/AC), and characterized for thermogravimetric analysis (TGA), Fourier-transformed infrared spectroscopy (FTIR), diffraction light scattering, Powder XRD, and scanning electron microscopy (SEM/EDX). The adsorption characteristics such as kinetics, mechanism and isotherms of toluene were studied. The kinetic data were analyzed using a pseudo-first-order, pseudo-second-order kinetic equation elovick and intraparticle diffusion model. The pseudo-first-order and pseudo-second-order kinetic model describes experimental data better than elovich and intraparticle kinetic model. The results show that Qtg-Ag/AC exhibits superior adsorption capacity (39.352 mg g−1) compared to AC (30.093 mg g−1). Langmuir, Freundlich and Tenkim model for AC and Qtg-Ag/AC described the equilibrium data better. In order to determine the best-fit kinetic model for each system, two error analysis methods, namely, chi-square (χ2) and residual mean squared error (RMSE) were used to evaluate the data. The outcomes of the study demonstrate that our modified (Qtg-Ag/AC) can be a good prototype for the adsorption of toluene from wastewater.

Utilizing of 500 MHz Ground Penetrating Radar Antenna Coupling at Different Interval Heights Clearance

Understanding the characteristics of the GPR antenna for underground utility detection purposes is critical to ensuring that buried pipes and cables can be accurately located. Considering the ground coupling antenna is the most effective way to carry out the GPR scan, most of the operators applied this setup to ensure the maximum amount of the electromagnetic wave could be penetrated into the ground. However, the contact between the GPR and the ground surface causes the bottom of the antenna enclosure to become scratched and worn out. Due to the cost and time involved in transportation to replace the antenna protector, a few GPR operators changed the protector to the cheapest one, which affects the characteristics of the antenna and could alter the depth of the underground utilities. A 500 MHz GPR antenna are tested to identify the optimum antenna coupling based on the accuracy assessment at various interval clearance heights in order to reduce damage and preserve the antenna. An experiment was carried out at Department of Survey and Mapping Malaysia (JUPEM) Kuala Lumpur’s Underground Utility Test Base. Based on the result of the signal strength, the higher the antenna is lifted, the higher the rate of the signal becomes weak or lost. For the depth accuracy test, the high differences in depth when the antenna is lifted up at 4cm to 6cm based on the result of comparison with the standard values provided by the JUPEM. On the other hand, the accuracy of each antenna height is compared and validated based on the Residual Mean Square Error (RMSE). Overall, this study could lead to the conclusion that the result of the assessment is within the acceptable RMSE tolerance, which is below 0.1m. The objective of the research is successfully achieved when the antenna at a 3cm to 6cm interval height setup provides the best clearance based on the criteria of signal strength, depth comparison, and accuracy assessment. Finally, the issue of contact between the antenna and the ground surface, which causes the antenna to become scratched and worn out, could be reduced and preserved as a result of the findings.

Deep Learning-Based Univariate Prediction of Daily Rainfall: Application to a Flood-Prone, Data-Deficient Country

There are several attempts to model rainfall time series which have been explored by members of the hydrological research communities. Rainfall, being one of the defining factors for a flooding event, is rarely modeled singularly in deep learning, as it is usually performed in multivariate analysis. This study will attempt to explore a time series modeling method in four subcatchments located in Samar, Philippines. In this study, the rainfall time series was treated as a signal and was reconstructed into a combination of a ‘smoothened’ or ‘denoised’ signal, and a ‘detailed’ or noise signal. The discrete wavelet transform (DWT) method was used as a reconstruction technique, in combination with the univariate long short-term memory (LSTM) network method. The combination of the two methods showed consistently high values of performance indicators, such as Nash–Sutcliffe efficiency (NSE), correlation coefficient (CC), Kling–Gupta efficiency (KGE), index of agreement (IA), and Legates–McCabe index (LMI), with mean average percentage error (MAPE) values at almost zero, and consistently low values for both residual mean square error (RMSE) and RMSE-observations standard deviation ratio (RSR). The authors believe that the proposed method can give efficient, time-bound results to flood-prone countries such as the Philippines, where hydrological data are deficient.

Application of inclusive multiple model for the prediction of saffron water footprint

Applying new approaches in the management of water resources is a vital issue, especially in arid and semi-arid regions. The water footprint is a key index in water management. Therefore, it is necessary to predict its changes for future durations. The soft computing model is one of the most widely used models in predicting and estimating agroclimatic variables. The purpose of this study is to predict the green and blue water footprints of saffron product using the soft computing model. In order to select the most effective variables in prediction water footprints, the individual input was eliminated one by one and the effect of each on the residual mean square error (RMSE) was measured. In the first stage, the Group Method of Data Handling (GMDH) and evolutionary algorithms have been applied. In the next stage, the output of individual models was incorporated into the Inclusive Multiple Model (IMM) as the input variables in order to predict the blue and green water footprints of saffron product in three homogenous agroclimatic regions. Finally, the uncertainty of the model caused by the input and parameters was evaluated. The contributions of this research are introducing optimized GMDH and new ensemble models for predicting BWF, and GWF, uncertainty analysis and investigating effective inputs on the GWF and BWF. The results indicated that the most important variables affecting green and blue water footprints are plant transpiration, evapotranspiration, and yield, since removing these variables significantly increased the RMSE (range=11–25). Among the GMDH models, the best performance belonged to NMRA (Naked Mole Ranked Algorithm) due to the fast convergence and high accuracy of the outputs. In this regard, the IMM has a better performance (FSD=0.76, NSE=0.95, MAE) = 8, PBIAS= 8) than the alternatives due to applying the outputs of several individual models and the lowest uncertainty based on the parameters and inputs of the model (p = 0.98, r = 0.08).

Predicting rejection of emerging contaminants through RO membrane filtration based on ANN-QSAR modeling approach: trends in molecular descriptors and structures towards rejections.

In this work, a quantitative structure-activity relationship (QSAR) study was performed on a set of emerging contaminants (ECs) to predict their rejections by reverse osmosis membrane (RO). A wide range of molecular descriptors was calculated by Dragon software for 72 ECs. The QSAR data was analyzed by an artificial neural network method (ANN), in which four out of 3000 theoretical molecular descriptors were chosen and their significance was computed based on the Garson method. The significance trends of descriptors were as follows in descending order: ESpm14u > R2e > SIC1 > EEig03d. The selected descriptors were ranked based on their importance and then an explorative study was conducted on the QSAR data to show the trends in molecular descriptors and structures toward the rejections values of ECs. The MLR algorithm was used to make a linear model and the results were compared with those of the nonlinear ANN algorithm. The comparison results revealed it is necessary to apply the ANN model to this data with non-linear properties. For the whole dataset, the correlation coefficient (R2) and residual mean squared error (RMSE) of the ANN and MLR methods were 0.9528, 6.4224; and 0.8753, 11.3400, respectively. The comparison results showed the superiority of ANN modeling in the analysis of ECs' QSAR data.

Kinetics and Thermodynamics of β-Carotene Adsorption onto Acid-Activated Clays Modified by Zero Valent Iron

The adsorption of β-carotene from crude palm oil onto acid-activated clay and clay modified by zero valent iron (ZVI) was investigated in this work. Spectroscopic studies including FTIR, XRD, and SEM were used for its characterization. The adsorption characteristics such as kinetics, mechanism, isotherms, and thermodynamics of β-carotene were studied. The kinetic data were analyzed using the pseudo-first-order kinetic equation, pseudo-second-order kinetic equation, and intraparticle diffusion model. The pseudo-second-order kinetic model is the only one that describes the experimental data well (R2 ≥ 0.969). The chemical analysis of bulk clay showed that the predominant oxides are Al2O3 (57.91 wt%), Fe2O3 (32.54 wt%), SiO2 (3.09 wt%), K2O (2.37 wt%), and CaO2 (1.73 wt%). The adsorption capacity increases with an increase in temperature. The equilibrium data were described better by the Freundlich model for all clays. To determine the best fit kinetic model for each system, three error analysis methods, namely, chi-square (χ2), residual mean squared error (RMSE), and mean percent error (%APE) were used to evaluate the data. A thermodynamic study demonstrated that β-carotene adsorption is spontaneous, endothermic, and an entropy driven process for both forms of clay.

Evaluation of Parameter Identification of a Real Manipulator Robot

Given the widespread use of the Kalman filter in robotics, an increasing number of researchers devote themselves to its study and application. This work underscores the importance of this filter while analyzing the modifications made to the same to improve its performance and reduce its deficiencies in some fields and presenting some of its applications in robotics. The following methods are presented in this study: least squares (LS), Hopfield Neural Networks (HNN), Extended Kalman filter (EKF), and Unscented Kalman filter (UKF). These methods are used in the parameter identification of a Selective Compliant Assembly Robot Arm (SCARA) robot with 3-Degrees of Freedom (3-DoF) and a clamp at its end. The dynamic model of this robot is obtained and employed to identify its parameters; then, the identification results are compared considering the difference between the obtained parameters and the real values of the robot parameters; in this comparison, the good results yielded by the LS and UKF method stand out. Subsequently, the obtained parameters through each method are validated by measuring different performance indexes—during trajectory tracking—such as: Residual Mean Square Error (RMSE), Integral of the Absolute Error (IAE), and the Integral of the Square Error (ISE). In this way, a comparison of the robot’s performance is possible.

Tree Height Derivation under Homogeneous Tree Pattern by Segmentation Algorithm

Tree height is an important element in plantation areas to determine the tree's maturity for harvesting. Airborne Laser Scanning (ALS) technology can give accurate elevation using point cloud data. However, this technology is quite expensive and unsuitable for small areas and low-budget projects. This research focuses on the UAV technology, exploring the appropriate methods of determining a tree height from the UAV’s photogrammetry images. This research aims to evaluate the tree height from the delineation of the tree crown. Three different algorithms have been used in this research to delineate tree crowns. The delineate tree crowns were extracted in vector format, and the crowns were used to calculate the tree height using the specific formula. The results of tree heights were assessed using Residual Mean Square Error (RMSE) to determine the accuracy of the outcome. It was found that the OBIA algorithm gives the best accuracy among these three algorithms. It is followed by WS algorithm and then IWS algorithm. The OBIA algorithm can give an accuracy of about 0.444m at 40m and 0.381m for 60 altitude. The accuracy for WS and OBIA stated that the 60m altitude gives the better accuracy compared to 40m altitude. However, the IWS gives the vice versa result. This research could help the planters to manage their plantations for the harvesting process.

A New Ensemble Index for Extracting Predictable Drought Features from Multiple Historical Simulations of Climate

Drought assessment and forecasting under an ensemble of multiple climate simulation models play important role in early warning drought mitigation policies. This research provides a new ensemble drought measure – the Multivariate Multi-Scaler Forecastable Standardized Drought Index (MMFSDI). At a particular georeferenced point, the MMFSDI uses time-series data of precipitation from multiple climate simulation models for the characterization of drought. The methodology of MMFSDI is mainly based on Forecastable Component Analysis (FCA) and K-Component Gaussian Mixture Distribution (K-CGMD). In application, historical simulated data of precipitation from 23 climate models of Coupled Model Intercomparison Project Phase 6 (CMIP6) at fifty grid points scattered over the Tibet Plateau region are considered to evaluate the applicability of MMFSDI. For comparative analysis, the forecasting performance of MMFSDI is compared with Standardized Precipitation Index (SPI) using Residual Mean Square Error (RMSE) and Mean Average Error (MAE) under Auto-Regressive Integrated Moving Average (ARIMA) and Artificial Neural Networks (ANN) models. Outcomes associated with this research shows that 1) the first component of FCA as an ensemble of multiple climate models is more describable than simple model averaging, 2) the strong consistency between MMFSDI and SPI makes MMFSDI as an alternative multi-model drought measure, and 3) the implications of ARIMA and ANN revealed that MMFSDI has inherited feature for forecasting drought. In summary, the finding of the research argues to substitute SPI with MMFSDI as MMFSDI has inherited forecasting ability.

Optimised point estimators for multi-stage single-arm phase II oncology trials

ABSTRACT The uniform minimum variance unbiased estimator (UMVUE) is, by definition, a solution to removing bias in estimation following a multi-stage single-arm trial with a primary dichotomous outcome. However, the UMVUE is known to have large residual mean squared error (RMSE). Therefore, we develop an optimisation approach to finding estimators with reduced RMSE for many response rates, which attain low bias. We demonstrate that careful choice of the optimisation parameters can lead to an estimator with often substantially reduced RMSE, without the introduction of appreciable bias.

Comparison of two software programs for fitting one- and two-compartment age-dependent non-linear digestion models for ruminants: empirical data

Context The total mass and kinetics of feed particles through the digestive tract affect feed intake, nutrient excretion and emissions by ruminants. Use of models to calculate digesta kinetics parameters will assist managers in mitigating enteric methane emission and producing food more sustainably. Aims We evaluated two software programs for fitting parameters to one- and two-compartment age-dependent digesta kinetic models from faecal-marker concentration datasets. Methods We examined biases (mean differences) and standard deviations (differences) of one-compartment (G2) and two-compartment (G2G1) models with a gamma-2 distribution in the age-dependent compartment when parameterised with two different software programs (R or SAS), using 41 datasets of ytterbium concentrations in faecal samples collected at discrete times. Faecal-marker concentration datasets were fitted to G2 and G2G1 models with each software program. The resulting model parameters, K0, λ or λ1, K2 and τ, were used to calculate the digesta kinetics parameters: particle passage rate, gastrointestinal dry matter fill, faecal dry matter output, gastrointestinal mean retention time and rumen retention time. We evaluated bias and standard deviation for model and digesta kinetic parameters across the entire range of average values, but also within low, medium and high percentile range-of-value subsets (5–35%, 35–65% and 65–95%) between software programs. Key results When datasets were fitted to the G2 model, all converged for both software programs, but when fitted to the G2G1 model by the SAS program, three observations did not converge. Bias and standard deviation of differences between software packages were small, but the G2G1 model produced smaller bias and standard deviation of differences. Bias and standard deviation of differences for digesta kinetics estimates across the percentile groups did not vary linearly for most model estimates and were small relative to the magnitude of the values. Conclusions Model parameters and digesta kinetics estimates derived from R and SAS software programs can be used interchangeably in nutritional modelling. Two-compartment models (G2G1) can be more problematic to fit, but residual mean-square errors are usually smaller. Implications Model parameters from both G2 and G2G1 models can be used to derive unbiased estimates of digesta kinetics from either R or SAS software program.

Forecasting the future number of pertussis cases using data from Google Trends

BackgroundAlternative methods could be used to enhance the monitoring and forecasting of re-emerging conditions such as pertussis. Here, whether data on the volume of Internet searching on pertussis could complement traditional modeling based solely on reported case numbers was assessed. MethodsSARIMA models were fitted to describe reported weekly pertussis case numbers over a four-year period in Germany. Pertussis-related Google Trends data (GTD) was added as an external regressor. Predictions were made by the models, both with and without GTD, and compared with values within the validation dataset over a one-year and for a two-weeks period. ResultsPredictions of the traditional model using solely reported case numbers resulted in an RMSE (residual mean squared error) of 192.65 and 207.8, a mean absolute percentage error (MAPE) of 58.59 and 72.1, and a mean absolute error (MAE) 169.53 and 190.53 for the one-year and for the two-weeks period, respectively. The GTD expanded model achieved better forecasting accuracy (RMSE: 144.22 and 201.78), a MAPE 43.86, and 68.54 and a MAE of 124.46 and 178.96. Corrected Akaike Information Criteria also favored the GTD expanded model (1750.98 vs. 1746.73). The difference between the predictive performances was significant when using a two-sided Diebold-Mariano test (DM value: 6.86, p < 0.001) for the one-year period. ConclusionInternet-based surveillance data enhanced the predictive ability of a traditionally based model and should be considered as a method to enhance future disease modeling.

Prediction of storage time in different seafood based on color values with artificial neural network modeling.

The determination of storage time in seafood could be performed by microbiological, chemical and sensory analysis. Among these mentioned methods color changes are one part of sensory analysis and are prior acceptance criteria from the point of consumers' view. In this study, a feedforward artificial neural network (ANN) model was developed to predict the storage time of seafood based on L*, a* and b* values. A total of 205 data set were compiled from the literature that represents the color changes of different seafood products to train and test the ANN model. Another set of data (n = 45) were used for the validation of developed ANN model. A multi-layer perceptron (MLP) was applied for the determination of agreements between input and output data. The most accurate topology were determined in accordance with the changes in the values of correlation coefficients (R2) and mean square errors (MSE) and found to be 30 neurons in the layer (R2 = 0.81 and MSE = 0.2). The performance of ANN model was evaluated based on 6 criteria such as Mean Absolute Deviation (MAD), Mean Square Errors (MSE), Residual Mean Square Errors (RMSE), Correlation Coefficient (R2), Mean Absolute Error (MAE) and F-test statistics and found to be 0.2, 0.05, 0.002, 0.8, 0.71 and 1.06, respectively. Moreover, predicted and observed storage time values were fitted and regression coefficient was found to be 0.85. In accordance with the results of this study, the proposed ANN model is accurate, reliable, and proper for the estimation of storage time in seafood products.

Metabolizable Protein: 1. Predicting Equations to Estimate Microbial Crude Protein Synthesis in Small Ruminants.

Microbial crude protein (MCP) produced in rumen could be estimated by a variety of protocols of experimental sampling and analysis. However, a model to estimate this value is necessary when protein requirements are calculated for small ruminants. This model could be useful to calculate rumen degradable protein (RDP) requirements from metabolizable protein (MP). Then, our objective was to investigate if there is a difference in MCP efficiency between sheep and goats, and to fit equations to predict ruminal MCP production from dietary energy intake. The database consisted of 19 studies with goats (n = 176) and sheep (n = 316), and the variables MCP synthesis (g/day), total digestible nutrients (TDN), and organic matter (OM) intakes (g/day), and OM digestibility (g/kg DM) were registered for both species. The database was used for two different purposes, where 70% of the values were sorted to fit equations, and 30% for validation. A meta-analytical procedure was carried out using the MIXED procedure of SAS, specie was considered as the fixed dummy effect, and the intercept and slope nested in the study were considered random effects. No effect of specie was observed for the estimation of MCP from TDN, digestible Organic Matter (dOM), or metabolizable energy (ME) intakes (P > 0.05), considering an equation with or without an intercept. Therefore, single models including both species at the same fitting were validated. The following equations MCP (g/day) = 12.7311 + 59.2956 × TDN intake (AIC = 3,004.6); MCP (g/day) = 15.7764 + 62.2612 × dOM intake (AIC = 2,755.1); and MCP (g/day) = 12.7311 + 15.3000 × ME intake (AIC = 3,007.3) presented lower values for the mean square error of prediction (MSEP) and its decomposition, and similar values for the concordance correlation coefficient (CCC) and for the residual mean square error (RMSE) when compared with equations fitted without an intercept. The intercept and slope pooled test was significant for equations without an intercept (P < 0.05), indicating that observed and predicted data differed. In contrast, predicted and observed data for complete equations were similar (P > 0.05).

Integrated meta-heuristics finite difference method for the dynamics of nonlinear unipolar electrohydrodynamic pump flow model

In this study, a novel design of integrated biological inspired computational heuristics is presented for the dynamics of nonlinear unipolar electrohydrodynamic (UP-EHD) pump flow model by exploiting the competency of finite difference method (FDM) for discretization, global search viability of genetic algorithms (GAs) and local search efficiency of active-set method (ASM), i.e., FDM-GA-ASM. The FDM is incorporated to transform the differential equations of the UP-EHD pump flow model into a system of nonlinear algebraic equations. The cost function is constructed through the mean-square residual error by mimicking forward, central and backward difference schemes viable for a broader range of physical models. The optimum solution is achieved by the integration of global search with GAs and local search of ASM for speedy refinements. The designed stochastic numerical solver FDM-GA-ASM investigates the critical physical parameters, i.e., charge density, electric field and electric potential by varying electrical slip, Reynolds number and source number of the UP-EHD model. Statistical observations in terms of probability plots, histogram illustrations, boxplots for the cost function, mean absolute error, root mean squared error and Nash–Sutcliffe efficiency metrics are used to validate the efficiency of the FDM-GA-ASM scheme for the three variants of the UP-EHD model. The designed FDM-GA-ASM is a promising numerical computing solver for nonlinear differential systems in engineering and technology.

An approach to quantify natural durability of Eucalyptus bosistoana by near infrared spectroscopy for genetic selection

Natural durability within a timber species can be variable. Hence efficient assessments of natural durability are required to ensure quality either through tree breeding or segregation during production. In this study, first the relationship between extractive content and mass loss of Eucalyptus bosistoana heartwood caused by a white and a brown-rot fungus was validated. Then the ability of NIR spectroscopy as a high-throughput method to evaluate heartwood decay resistance was examined. Finally the NIR method was applied to a tree breeding trial. A correlation between extractive content and mass loss against the white-rot fungus (Perenniporia tephropora) and the brown-rot fungus (Coniophora olivacea) were found. Analysis of NIR spectra indicated that this relationship is causal with shared bands for mass loss and extractive content models at 6650, 6017, 5265 and 4659 cm−1. Partial least squares regression (PLSR), supplemented with spectra normalisation and variable selection, allowed prediction of mass loss with a residual mean square error (RMSE) of 7.48 % and 5.76 % for the white-rot and brown-rot, respectively. This level of precision allowed the characterisation of a E. bosistoana resource which showed a range of mass loss from 0 to 60 %. Genetic control was found for mass loss by the white-rot ( h2 = 0.70 and 0.24) and the brown-rot ( h2 = 0.15 and 0.13) at two sites in New Zealand. The rankings were correlated between sites, with genetic correlations (Rg2) of 0.69 and 0.63 for white-rot and brown-rot, respectively, as well as to the predicted extractive content (0.82 to 0.92). However, the study indicated a significant site effect on the decay resistance of the E. bosistoana heartwood. In summary, this study has shown that the decay resistance could be assessed rapidly and efficiently using NIR technology for genetic selection.

Developing Children's Oral Health Assessment Toolkits Using Machine Learning Algorithm.

Evaluating children's oral health status and treatment needs is challenging. We aim to build oral health assessment toolkits to predict Children's Oral Health Status Index (COHSI) score and referral for treatment needs (RFTN) of oral health. Parent and Child toolkits consist of short-form survey items (12 for children and 8 for parents) with and without children's demographic information (7 questions) to predict the child's oral health status and need for treatment. Data were collected from 12 dental practices in Los Angeles County from 2015 to 2016. We predicted COHSI score and RFTN using random Bootstrap samples with manually introduced Gaussian noise together with machine learning algorithms, such as Extreme Gradient Boosting and Naive Bayesian algorithms (using R). The toolkits predicted the probability of treatment needs and the COHSI score with percentile (ranking). The performance of the toolkits was evaluated internally and externally by residual mean square error (RMSE), correlation, sensitivity and specificity. The toolkits were developed based on survey responses from 545 families with children aged 2 to 17 y. The sensitivity and specificity for predicting RFTN were 93% and 49% respectively with the external data. The correlation(s) between predicted and clinically determined COHSI was 0.88 (and 0.91 for its percentile). The RMSEs of the COHSI toolkit were 4.2 for COHSI (and 1.3 for its percentile). Survey responses from children and their parents/guardians are predictive for clinical outcomes. The toolkits can be used by oral health programs at baseline among school populations. The toolkits can also be used to quantify differences between pre- and post-dental care program implementation. The toolkits' predicted oral health scores can be used to stratify samples in oral health research. This study creates the oral health toolkits that combine self- and proxy- reported short forms with children's demographic characteristics to predict children's oral health and treatment needs using Machine Learning algorithms. The toolkits can be used by oral health programs at baseline among school populations to quantify differences between pre and post dental care program implementation. The toolkits can also be used to stratify samples according to the treatment needs and oral health status.

Predicting tick-borne encephalitis using Google Trends

Data generated through public Internet searching offers a promising alternative source of information for monitoring and forecasting of infectious disease. Here future cases of tick-borne encephalitis (TBE) were predicted using traditional weekly case reports, both with and without Google Trends data (GTD). Data on the weekly number of acute, confirmed TBE cases in Germany were obtained from the Robert Koch Institute. Data relating to the volume of Internet searching on TBE was downloaded from the Google Trends website. Data were split into training and validation parts. A SARIMA (0,1,1) (1,1,1) [52] model was used to describe the weekly TBE case number time series. Google Trends Data was used as an external regressor in a second, as optimal identified SARIMA (4,1,1) (1,1,1) [52] model. Predictions for the number of future cases were made with both models and compared with the validation dataset. GTD showed a significant correlation with reported weekly case numbers of TBE (p < 0.0001). A comparison of forecasted values with reported ones resulted in an RMSE (residual mean squared error) of 0.71 for the model without Google search values, and an RMSE of 0.70 for the Google Trends values enhanced model. However, difference between predictive performances was not significant (Diebold Mariano test, p-value = 0.14).

Calibration of near infrared spectroscopy (NIRS) data of three Eucalyptus species with extractive contents determined by ASE extraction for rapid identification of species and high extractive contents

Abstract Plantations of naturally durable timber species could substitute unsustainably harvested wood from tropical forests or wood treated with toxic preservatives. The New Zealand Dryland Forests Initiative (NZDFI) has established a tree-breeding program to develop genetically improved planting stock for durable eucalyptus plantations. In this study the durable heartwood of Eucalyptus bosistoana, Eucalyptus globoidea and Eucalyptus argophloia was characterized by near infrared (NIR) spectroscopy and NIR data was calibrated with the extractives content (EC), determined by accelerated solvent extraction (ASE) extraction, by means of a partial least squares regression (PLSR) model. It was possible to predict the EC content in the range of 0.34–18.9% with a residual mean square error (RMSE) of 0.9%. Moreover, the three species could also be differentiated by NIR spectroscopy with 100% accuracy, i.e. NIR spectroscopy is able to segregate timbers from mixed species forest plantations.