Root Mean Square Error Of Validation Research Articles

Rapid determination of chicken meat ratios in Beef Mixtures and Beef Sausages by Near Infrared Reflectance (NIR) spectroscopy

This study aims to determine the percentage of chicken meat in beef and chicken mixtures, which is the most common form of beef adulteration. Ground beef and beef sausages were prepared with mixtures containing chicken meat, ranging from 0.0% to 100.0% with 5.0% increments, and analyzed using a near-infrared spectroscopy device. Optimal analysis conditions were determined through the examination of a wide range of regression models. The best regression model for ground beef mixtures yielded the following results: RMSEC (Root Mean Square Error of Calibration): 2.35, RMSEV (Root Mean Square Error of Validation): 3.36, R2C (R-Value Calibration): 0.99, R2V (R-Value Validation): 0.98. The results for beef sausages were as follows: RMSEC: 2.56, RMSEV: 3.66, R2C: 0.99, R2V: 0.98. As a result, the chicken meat content in beef mixtures was detected with a margin of error of 2.05%, while the chicken meat content in beef sausages was detected with a margin of error of 2.12%.

Estimating body weight of caged sea cucumbers (Apostichopus japonicus) using an underwater time-lapse camera and image analysis by semantic segmentation

Image analysis is being developed to improve the efficiency of fishery and aquaculture technologies. Optical cameras are an easy and cost-effective method for monitoring fish and other species. In this study, a monitoring system that combines an underwater time-lapse camera and a deep learning-based image analysis was developed for utilization in integrated multi-trophic aquaculture (IMTA). The sea cucumber (Apostichopus japonicus) was used as a target species because the technology necessary for estimating growth, particularly in terms of weight, of caged sea cucumber using an underwater environment is still under study. Therefore, semantic segmentation was applied to classify the images into caged sea cucumbers and various underwater backgrounds. Multiple images of sea cucumbers were captured in a water tank that mimicked the box cage used in IMTA, and their body weights were measured simultaneously. For model development, approximately 1,300 images were prepared for the training and validation processes. The model then achieved an IoU (Intersection over Union) of approximately 94% for the validation data. Next, the pixel numbers of sea cucumbers were converted into an area calculated using the size of the cage net as the background. The relationship between the area and weight of sea cucumbers yielded an approximate line for estimating body weight. As a result, the approximation line had a coefficient of determination of R2 = 0.87 for training and validation data and RMSE (Root Mean Square Error) =1.81 and 6.78 g for sea cucumbers less than 10 and 110 g, respectively. Using the model, test images in an actual IMTA situation were applied, and the estimated body weights were close to the measured values for small sea cucumbers. If we apply this model to images obtained over an extended period, the growth of sea cucumbers in a time series can be understood.

Where will the next oil spill incident in the Niger Delta region of Nigeria occur?

Oil spill incidents are almost a daily occurrence within the Niger Delta region of Nigeria with far-reaching environmental, economic and social consequences. This study aimed at understanding the spatial and temporal context of the problem as a panacea for forecasting likely locations of oil spill incidents within the region. About 76.77% of crude oil spilt in the Niger Delta is lost to the environment with only about 23% of the crude oil recovered from the environment. This represents a very worrying statistic in terms of the known and unknown negative impacts of oil spills. Space Time Pattern Mining (STPM) tools were adapted to explore and interrogate historical spill data. Time series forecasting was then used for forecasting possible locations of future oil spills within the region. Results show that there is a pattern of oil spill occurrences in the Niger Delta with statistically significant hotspots identified in Rivers State, Bayelsa State and Delta State. The forecast root mean square error (RMSE) and forecast validation RMSE are −1.016328 and 1.035992 respectively. This suggests an ability of the model to fairly predict likely locations of future oil spills. This was further verified by counting the number of spills that occur within any area based on the predicted likelihood of spill occurrence. This study has shown that STPM tools can be deployed to understand the occurrence and prediction of oil spill incidents. This will ultimately aid in the deployment of scarce management resources to where they are most needed.

Estimating the SPAD of Litchi in the Growth Period and Autumn Shoot Period Based on UAV Multi-Spectrum

The relative content of chlorophyll, assessed through the soil and plant analyzer development (SPAD), serves as a reliable indicator reflecting crop photosynthesis and the nutritional status during crop growth and development. In this study, we employed machine learning methods utilizing unmanned aerial vehicle (UAV) multi-spectrum remote sensing to predict the SPAD value of litchi fruit. Input features consisted of various vegetation indices and texture features during distinct growth periods, and to streamline the feature set, the full subset regression algorithm was applied for dimensionality reduction. Our findings revealed the superiority of stacking models over individual models. During the litchi fruit development period, the stacking model, incorporating vegetation indices and texture features, demonstrated a validation set coefficient of determination (R2) of 0.94, a root mean square error (RMSE) of 2.4, and a relative percent deviation (RPD) of 3.0. Similarly, in the combined litchi growing period and autumn shoot period, the optimal model for estimating litchi SPAD was the stacking model based on vegetation indices and texture features, yielding a validation set R2, RMSE, and RPD of 0.84, 3.9, and 1.9, respectively. This study furnishes data support for the precise estimation of litchi SPAD across different periods through varied combinations of independent variables.

Multi-Source Soil Moisture Data Fusion Based on Spherical Cap Harmonic Analysis and Helmert Variance Component Estimation in the Western U.S.

Soil moisture (SM) is a vital climate variable in the interaction process between the Earth's atmosphere and land. However, global soil moisture products from various satellite missions and land surface models are affected by inherently discontinuous observations and coarse spatial resolution, which limits their application at fine spatial scales. To address this problem, this paper integrates three diverse types of datasets from in situ, satellites, and models through Spherical cap harmonic analysis (SCHA) and Helmert variance component estimation (HVCE) to produce 1 km of spatio-temporally continuous SM products with high accuracy. First, this paper eliminates the bias between different datasets and in situ sites and resamples the datasets before data fusion. Then, multi-source SM data fusion is performed based on the SCHA and HVCE methods. Finally, this paper evaluates the fused products from three aspects, including the performance of representative sites under different climate types, the overall performance of validation sites, and the comparison with other products. The results show that the fused products have better performance than other SM products. In the representative sites, the minimal correlation coefficient (R) of the fused products is above 0.85, and the largest root mean square error (RMSE) is below 0.040 m3 m-3. For all validation sites, the R and RMSE of the fused products are 0.889 and 0.036 m3 m-3, respectively, while the R for other products is below 0.75 and the RMSE is above 0.06 m3 m-3. In comparison to other SM products, the fused products exhibit superior performance, generally align more closely with in situ measurements, and possess the ability to accurately and finely capture the spatial and temporal variability of surface SM.

The use of non-invasive fluorescence spectroscopy to quantify phenolic content under red wine real-time fermentation conditions

Phenolic compounds play important roles in wine quality attributes such as colour, mouthfeel and ageing potential. The ability to monitor their extraction and implement appropriate vinification techniques relies on accurate phenolic analysis methods. Front-face fluorescence spectroscopy presents itself as a user-friendly, rapid and cost-effective alternative to other spectrophotometric methods. The main aim of the study was therefore to investigate the potential of fluorescence spectroscopy to directly measure phenolic content of red wine samples throughout red wine fermentations. Cabernet Sauvignon fermentations were monitored using fluorescence spectroscopy and UV–Visible spectrophotometry. Fermentation conditions were explored for their influence on the prediction accuracy of fluorescence-based regression models. The coefficient of correlation (R2val) and root mean square errors of validation (RMSEV) for models built using non-invasive spectral data obtained from unaltered samples were 0.96, 0.94 and 0.89 and 1.38 index units, 42.84 mg/L and 20.53 mg/L for total phenols (index units (IU)), total condensed tannins (mg/L), total anthocyanins (mg/L), respectively. Overall, the ability to obtain high quality spectra from unaltered samples simulating direct measurements from the fermentation vessel was demonstrated and holds potential for on-line automated systems or portable device applications.

Construction of a hyperspectral estimation model for total nitrogen content in Shajiang black soil

AbstractBackgroundThe real‐time non‐destructive estimation of soil total nitrogen content via hyperspectral remote sensing is important for crop fertilization and precision agriculture development.AimsThe study aimed to use hyperspectral technology for the construction of an estimation model using different methods for assessing total nitrogen content in Shajiang black soil.MethodsIn our study, soil samples were obtained from Shangshui County, Henan Province, China. Using hyperspectral data of soil samples, the original spectral reflectance was transformed into logarithmic, reciprocal, and first derivative spectra. Hyperspectral estimation models for assessing the total nitrogen content were built with spectral indices, competitive adaptive reweighted sampling (CARS) algorithms, and a combination of machine learning methods, including the partial least square regression, support vector machine (SVM), and random forest methods.ResultsThe results showed that after the first derivative transformation, the performance of normalized indices constructed with a combination of two bands at 1401 and 776 nm was better. The coefficient of determination (R2), root mean square error (RMSE), ratio of performance to deviation (RPD), and ratio of performance to the inter‐quartile range (RPIQ) values of the model were 0.84, 0.10 g kg–1, 2.25, and 4.04, respectively. Validation was performed using data independent of the modeling samples, and validation R2, RMSE, RPD, and RPIQ values were 0.92, 0.07 g kg–1, 2.56, and 4.19, respectively. The same sample was analyzed using a CARS algorithm for screening feature bands, and the 61 reciprocal reflectance bands selected as the input for the SVM exhibited the best performance. A modeling R2 value of 0.92, RMSE of 0.07 g kg–1, RPD of 3.09, RPIQ of 5.97, and validation R2 of 0.96, RMSE of 0.06 g kg–1, RPD of 3.24, and RPIQ of 8.47 were observed.ConclusionsThe estimation models constructed using the two methods exhibited good ability to determine the total nitrogen content. The accuracy of the SVM model was slightly higher than that of the indices model, but both models could facilitate the rapid estimation of total nitrogen content in Shajiang black soil. The findings can provide a technical reference for the estimation of the levels of soil total nitrogen and other nutrients.

Performance Evaluation of Hydroponic Wastewater Treatment Plant Integrated with Ensemble Learning Techniques: A Feature Selection Approach

Wastewater treatment and reuse are being regarded as the most effective strategy for combating water scarcity threats. This study examined and reported the applications of the Internet of Things (IoT) and artificial intelligence in the phytoremediation of wastewater using Salvinia molesta plants. Water quality (WQ) indicators (total dissolved solids (TDS), temperature, oxidation-reduction potential (ORP), and turbidity) of the S. molesta treatment system at a retention time of 24 h were measured using an Arduino IoT device. Finally, four machine learning tools (ML) were employed in modeling and evaluating the predicted concentration of the total dissolved solids after treatment (TDSt) of the water samples. Additionally, three nonlinear error ensemble methods were used to enhance the prediction accuracy of the TDSt models. The outcome obtained from the modeling and prediction of the TDSt depicted that the best results were observed at SVM-M1 with 0.9999, 0.0139, 1.0000, and 0.1177 for R2, MSE, R, and RMSE, respectively, at the training stage. While at the validation stage, the R2, MSE, R, and RMSE were recorded as 0.9986, 0.0356, 0.993, and 0.1887, respectively. Furthermore, the error ensemble techniques employed significantly outperformed the single models in terms of mean square error (MSE) and root mean square error (RMSE) for both training and validation, with 0.0014 and 0.0379, respectively.

Estimating the Texture of Purple Soils Using Vis-NIR Spectroscopy and Optimized Conversion Models

Soil texture is an indicator of soil physical structure which delivers many ecological functions of soils such as thermal regime, plant growth, and soil quality. However, traditional methods for soil texture measurement are time-consuming and labor-intensive. This study attempts to explore an indirect method for rapid estimating the texture of three subgroups of purple soils (i.e. calcareous, neutral, and acidic). 190 topsoil (0 - 10 cm) samples were collected from sloping croplands in Tongnan and Beibei Districts of Chongqing Municipality in China. Vis-NIR spectrum was measured and processed, and stepwise multiple linear regression (SMLR), partial least squares regression (PLSR), and back propagation neural network (BPNN) models were constructed to inform the soil texture. The clay fractions ranged from 4.40% to 27.12% while sand fractions ranged from 0.34% to 36.57%, hereby soil samples encompass three textural classes (i.e. silt, silt loam, and silty clay loam). For the original spectrum, the texture of calcareous and neutral purple soils was not significantly correlated with spectral reflectance and linear models (SMLR and PLSR) exhibited low prediction accuracy. The correlation coefficients and the goodness-of-fits between soil texture and the transformed spectra of all soil groups increased by continuum-removal (CR), first-order differential (R'), and second-order differential (R") transformations. Among them, the R" had the best performance in terms of improving the correlation coefficients and the goodness-of-fits. For the calcareous purple soil, the SMLR exceeds PLSR and BPNN with a higher coefficient of determination (R2) and the ratio of performance to inter-quartile distance (RPIQ) values and lower root mean square error of validation (RMSEV), but for the neutral and acidic purple soils, the PLSR model has a better prediction accuracy. In summary, the linear methods (SMLR and PLSR) are more reliable in estimating the texture of the three purple soil groups when using Vis-NIR spectroscopy inversion.

Sentinel-2 Satellite Imagery Application to Monitor Soil Salinity and Calcium Carbonate Contents in Agricultural Fields

The estuary tides affect groundwater dynamics; these areas are susceptible to waterlogging and salinity issues. A study was conducted on two fields with a total area of 60 hectares under a center pivot irrigation system that works with solar energy and belong to a commercial farm located in Northern Sudan. To monitor soil salinity and calcium carbonate in the area and stop future degradation of soil resources, easy, non-intrusive, and practical procedures are required. The objective of this study was to use remote sensing-determined Sentinel-2 satellite imagery using various soil indices to develop prediction models for the estimation of soil electrical conductivity (EC) and soil calcium carbonate (CaCO3). Geo-referenced soil samples were collected from 72 locations and analyzed in the laboratory for soil EC and CaCO3. The electrical conductivity of the soil saturation paste extract was represented by average values in soil dataset samples from two fields collected from the topsoil layer (0 to 15 cm) characteristic of the local salinity gradient. The various soil indices, used in this study, were calculated from the Sentinel-2 satellite imagery. The prediction was determined using the root mean square error (RMSE) and cross validation was done using coefficient of determination. The results of regression analysis showed linear relationships with significant correlation between the EC analyzed in laboratory and the salinity index-2 “SI2” (Model-1: R2 = 0.59, p = 0.00019 and root mean square error (RMSE = 1.32%) and the bare soil index “BSI” (Model-2: R2 = 0.63, p = 0.00012 and RMSE = 6.42%). Model-1 demonstrated the best model for predicting soil EC, and validation R2 and RMSE values of 0.48% and 1.32%, respectively. The regression analysis results for soil CaCO3 determination showed linear relationships with data obtained in laboratory and the bare soil index “BSI” (Model- 3: R2 = 0. 45, p = 0.00021 and RMSE = 1.29%) and the bare soil index “BSI” & Normalized difference salinity index “NDSI” (Model-4: R2 = 0.53, p = 0.00015 and RMSE = 1.55%). The validation confirmed the Model-3 results for prediction of soil CaCO3 with R2 and RMSE values of 0.478% and 1.29%, respectively. Future soil monitoring programs might consider the use of remote sensing data for assessing soil salinity and CaCO3 using soil indices results generated from satellite image (i.e., Sentinel-2).

Inversion of soil salinity in China's Yellow River Delta using unmanned aerial vehicle multispectral technique.

The rapid and accurate acquisition of soil property information, especially the soil salinity (SS), is required for saline soil management in the Yellow River Delta (YRD). In this study, Lijin County and Kenli District were selected as study area. Unmanned aerial vehicle (UAV) multispectral data and soil sample data were acquired from March 25 to 28, 2019. Pearson correlation and gray correlation analyses were first used to screen sensitive spectral bands/indices, which were used for model parameters construction. Three machine learning and one statistical analysis methods were used to construct the SS inversion models. The results found that the gray correlation coefficient value were greater than the Pearson coefficient value for all bands and indices. Based on the gray correlation coefficient, nine sensitive bands and indices were selected to construct 18 model parameters. By comparing the 4 models, it was concluded that the BPNN model had the highest inversion accuracy, and its calibration coefficient of determination (R2) and root mean square error (RMSE) were 0.769 and 1.342, respectively. The validation R2 and RMSE were 0.774 and 1.975, respectively, and the relative prediction deviation (RPD) was 2.963. The SS estimation results based on BPNN model were consistent with those of the field investigation. Rapid and accurate inversion of SS based on UAV multispectral technique was achieved in this study, which provides technical support for regional management.

Prediction of Apple Hybrid Offspring Aroma Based on Hyperspectral.

Used Random forest algorithm to construct a prediction model of aroma components based on the hybrid offspring of 'Honeycrisp' × 'Maodi', and different preprocessing methods were tried (Standardization (SS), First-order Derivative (D1) and Standard normal variate (SNV)). The aroma composition and content were determined by gas chromatography-mass spectrometry (GC-MS), and the main aroma components of apples were classified according to compound categories, including ester, aldehyde, ketone, alcohol. Taking the chemical groups as the research objects, the characteristic wavelengths were selected by grid search algorithm, and the characteristic wavelength-aroma chemical group model was established, and the same method was used to construct the model for single aroma components. The results show: SNV has the best noise removal effect among the five preprocessing methods. Under the SNV treatment, aroma chemical groups of apples showed a good correlation with the spectrum. The number of characteristic spectra of ester are 413, 493, 512, 551, 592, 600, 721, 727, 729, 733 nm, all in the visible light range. The determination coefficient (R2), the root mean square error (RMSE) and the ratio of the standard deviation values (RPD) of validation were 0.90, 4936.16 and 1.13. The characteristic spectrum of alcohols is 519, 562, 570, 571, 660, 676, 737, 738 nm, the range is close to that of ester. The R2 and RMSE of alcohol validation are 0.92 and 83.21, and RPD is 1.30. The number of characteristic spectra of aldehyde is 20, and the most important band is 1000 nm, which is outside the visible light range. The number of characteristic spectra of ketone is 15, and also has some distribution outside the visible light range. The R2 of aldehyde and ketone validation are 0.84 and 0.86. Except for cyclooctanol, the R2 of single aroma compound prediction model performed poorly. Based on the models, we tried to visualize alcohol, which can roughly represent their distribution on apple. Their distributions all show significant differences in the center and edge of apple, but the results are still rough due to the accuracy of models. In conclusion, the study can preliminarily prove that hyperspectral imaging technology (HSI) can perform non-destructive detection of aroma in apple hybrid offspring.

A CNN Regression Approach for Real-Time Estimation of Dielectrophoretic Force

This research presents a supervised learning framework with a Convolutional Neural Network (CNN) to precisely quantify dielectrophoretic force () in microfluidic devices using a regression approach. As opposed to using a classification paradigm in our prior work, here we recast the estimation task as a regression problem. Micrographs of pearl-chain arrangement in a DEP sensing device were processed using a using modified CNN architectures of AlexNet, ResNet-50, MobileNetV2, and GoogLeNet in order to predict the induced on yeast cells and polystyrene microbeads. Micrographs obtained from our dielectrophoresis (DEP) experiments with varied input voltages were preprocessed and used in building these deep regression models. prediction accuracy of the models was tested for using Mean Absolute Error (MAE), Mean Absolute Relative (MRE), Mean Squared Error (MSE), R-squared, and Root Mean Square Error (RMSE) as evaluation metrics. The results from the experiments show that the performance of our regression models had better prediction accuracy and generalization ability when compared to our classification models. ResNet-50 with RMSPROP gave the best performance, with a validation RMSE of 0.0918 on yeast cells while AlexNet with ADAM optimizer gave the best performance, with a validation RMSE of 0.1745 on microbeads. This provides a baseline for further studies in the application of deep learning and dielectrophoresis for DEP force estimation. Figure 1

Study on hyperspectral monitoring model of soil total nitrogen content based on fractional-order derivative

Realizing the real-time nondestructive monitoring of soil total nitrogen (STN) content is an important task to promote precision agriculture development. In this study, a water regulation experiment was performed. Winter wheat and summer soybean soils were taken as the research objects. The effects of fractional-order derivatives (FOD) on soil spectral reflectance were studied, and the hyperspectral monitoring models of STN content were constructed based on the full spectrum band and the uninformative variable elimination (UVE) -extracted band. The main results were as follows: The FOD showed a gradual change in the integer-order derivative (IOD). When calculating the FOD, the more fitting points there were, the smaller the FOD spectrum, and the higher the correlation with the STN content. In the range of orders 0–1, the correlation increased with increasing order. In the range of orders 1–2, the correlation first decreased and then increased. When constructing the model, stepwise multiple linear regression (SMLR) was used, but the accuracies of the SMLR models were low. Afterward, the bands were extracted by UVE, and the models were constructed by partial least squares regression (PLSR). The number of bands extracted by UVE ranged from 11 to 534, mainly in the range of 1000–2450 nm. Under the same conditions, the accuracy of the PLSR model constructed with the UVE-extracted bands was higher than that of the model constructed with the full band spectrum. Among all PLSR models, when the number of fitting points, the order, and the number of bands were 40, 0.67, and 272, respectively, the model accuracy was the highest. The coefficient of determination of cross-validation (R2cv), root mean square error of cross-validation (RMSEcv), coefficient of determination of validation (R2v), root mean square error of validation (RMSEv), and relative analysis error of validation (RPDv) of the model were 0.7317, 0.2244 g·kg−1, 0.7937, 0.1976 g·kg−1, and 2.1904, respectively. This model could realize the hyperspectral monitoring of STN content in wheat and soybean fields. In this study, hyperspectral spectroscopy was used to monitor the STN content, which provided a theoretical basis and technical support for the observations.

Comparing machine learning and deep learning regression frameworks for accurate prediction of dielectrophoretic force

An intelligent sensing framework using Machine Learning (ML) and Deep Learning (DL) architectures to precisely quantify dielectrophoretic force invoked on microparticles in a textile electrode-based DEP sensing device is reported. The prediction accuracy and generalization ability of the framework was validated using experimental results. Images of pearl chain alignment at varying input voltages were used to build deep regression models using modified ML and CNN architectures that can correlate pearl chain alignment patterns of Saccharomyces cerevisiae(yeast) cells and polystyrene microbeads to DEP force. Various ML models such as K-Nearest Neighbor, Support Vector Machine, Random Forest, Neural Networks, and Linear Regression along with DL models such as Convolutional Neural Network (CNN) architectures of AlexNet, ResNet-50, MobileNetV2, and GoogLeNet have been analyzed in order to build an effective regression framework to estimate the force induced on yeast cells and microbeads. The efficiencies of the models were evaluated using Mean Absolute Error, Mean Absolute Relative, Mean Squared Error, R-squared, and Root Mean Square Error (RMSE) as evaluation metrics. ResNet-50 with RMSPROP gave the best performance, with a validation RMSE of 0.0918 on yeast cells while AlexNet with ADAM optimizer gave the best performance, with a validation RMSE of 0.1745 on microbeads. This provides a baseline for further studies in the application of deep learning in DEP aided Lab-on-Chip devices.

Predictive formulas for different measures of cheese yield using milk composition from individual goat samples

The objective of this study was to develop formulas based on milk composition of individual goat samples for predicting cheese yield (%CY) traits (fresh curd, milk solids, and water retained in the curd). The specific aims were to assess and quantify (1) the contribution of major milk components (fat, protein, and casein) and udder health indicators (lactose, somatic cell count, pH, and bacterial count) on %CY traits (fresh curd, milk solids, and water retained in the curd); (2) the cheese-making method; and (3) goat breed effects on prediction accuracy of the %CY formulas. The %CY traits were analyzed in duplicate from 600 goats, using an individual laboratory cheese-making procedure (9-MilCA method; 9 mL of milk per observation) for a total of 1,200 observations. Goats were reared in 36 herds and belonged to 6 breeds (Saanen, Murciano-Granadina, Camosciata delle Alpi, Maltese, Sarda, and Sarda Primitiva). Fresh %CY (%CYCURD), total solids (%CYSOLIDS), and water retained (%CYWATER) in the curd were used as response variables. Single and multiple linear regression models were tested via different combinations of standard milk components (fat, protein, casein) and indirect udder health indicators (UHI; lactose, somatic cell count, pH, and bacterial count). The 2 %CY observations within animal were averaged, and a cross-validation (CrV) scheme was adopted, in which 80% of observations were randomly assigned to the calibration (CAL) set and 20% to the validation (VAL) set. The procedure was repeated 10 times to account for sampling variability. Further, the model presenting the best prediction accuracy in CrV (i.e., comprehensive formula) was used in a secondary analysis to assess the accuracy of the %CY predictive formulas as part of the laboratory cheese-making procedure (within-animal validation, WAV), in which the first %CY observation within animal was assigned to CAL, and the second to the VAL set. Finally, a stratified CrV (SCrV) was adopted to assess the %CY traits prediction accuracy across goat breeds, again using the best model, in which 5 breeds were included in CAL and the remaining one in the VAL set. Fitting statistics of the formulas were assessed by coefficient of determination of validation (R2VAL) and the root mean square error of validation (RMSEVAL). In CrV, the formula with the best prediction accuracy for all %CY traits included fat, casein, and UHI (R2VAL = 0.65, 0.96, and 0.23 for %CYCURD, %CYSOLIDS, and %CYWATER, respectively). The WAV procedure showed R2VAL higher than those obtained in CrV, evidencing a low effect of the 9-MilCA method and, indirectly, its high repeatability. In the SCrV, large differences for %CYCURD and %CYWATER among breeds evidenced that the breed is a fundamental factor to consider in %CY predictive formulas. These results may be useful to monitor milk composition and quantify the influence of milk traits in the composite selection indices of specific breeds, and for the direct genetic improvement of cheese production.

Estimating global downward shortwave radiation from VIIRS data using a transfer-learning neural network

In recent years, machine learning (ML) has been successfully used in estimating downward shortwave radiation (DSR). To achieve global estimations, traditional ML models need sufficient ground measurements covering various atmospheric and surface conditions globally, which is difficult to accomplish. Training on the simulated data of a radiative transfer model (RTM) is a possible solution, but widely used RTMs ignore some complex cloud conditions which brings bias to simulations. In this study, a neural network applied with the transfer-learning (TL) concept is introduced to utilize both radiative transfer simulations and ground measurement data, achieving global DSR estimation with only top-of-atmosphere and surface albedo at local solar noon as inputs. The proposed method estimates both instantaneous and daily DSR from Visible Infrared Imaging Radiometer Suite (VIIRS) data at 750-m resolution, and both the estimates are validated by 40 independent stations globally. The root mean-square error and relative root mean square error of instantaneous DSR validation over 25 Baseline Surface Radiation Network, seven Surface Radiation Network, and eight Greenland Climate Network stations in 2013 were 91.2 (16.1%), 106.3 (18.3%), 75.0 (24.2%) W/m2, respectively, and the daily validation achieved 30.8 (15.5%), 33.5 (17.6%), and 31.3 (14.4) W/m2, respectively. The proposed method presents significant high accuracy over polar regions and similar performances over other areas compared with traditional ML models, physics models (e.g., look-up tables and direct estimations), and existing DSR products. The algorithm is also applied to VIIRS swath data to test its global efficacy. Instantaneous mapping captures the spatial pattern of the cloud-mask product, and daily mapping shows spatial patterns similar to the Clouds and the Earth's Radiant Energy System Synoptic TOA and surface fluxes and clouds product, but with more detail. Further analysis indicates that model performance is less sensitive to the quantity of training data after TL has been incorporated. This study demonstrates the advantages of TL on boosting both the generality and accuracy of DSR estimation, which can potentially be applied to other variable retrievals.

Precise Monitoring of Soil Salinity in China's Yellow River Delta Using UAV-Borne Multispectral Imagery and a Soil Salinity Retrieval Index.

Monitoring salinity information of salinized soil efficiently and precisely using the unmanned aerial vehicle (UAV) is critical for the rational use and sustainable development of arable land resources. The sensitive parameter and a precise retrieval method of soil salinity, however, remain unknown. This study strived to explore the sensitive parameter and construct an optimal method for retrieving soil salinity. The UAV-borne multispectral image in China’s Yellow River Delta was acquired to extract band reflectance, compute vegetation indexes and soil salinity indexes. Soil samples collected from 120 different study sites were used for laboratory salt content measurements. Grey correlation analysis and Pearson correlation coefficient methods were employed to screen sensitive band reflectance and indexes. A new soil salinity retrieval index (SSRI) was then proposed based on the screened sensitive reflectance. The Partial Least Squares Regression (PLSR), Multivariable Linear Regression (MLR), Back Propagation Neural Network (BPNN), Support Vector Machine (SVM), and Random Forest (RF) methods were employed to construct retrieval models based on the sensitive indexes. The results found that green, red, and near-infrared (NIR) bands were sensitive to soil salinity, which can be used to build SSRI. The SSRI-based RF method was the optimal method for accurately retrieving the soil salinity. Its modeling determination coefficient (R2) and Root Mean Square Error (RMSE) were 0.724 and 1.764, respectively; and the validation R2, RMSE, and Residual Predictive Deviation (RPD) were 0.745, 1.879, and 2.211.

Applicability of Near Infrared Reflectance Spectroscopy to Predict Amylose Contents of Single-Grain Maize

Near infrared reflectance spectroscopy (NIRS) and reference data were used to determine the amylose contents of single maize seeds to enable rapid, effective selection of individual seeds with desired traits. To predict the amylose contents of a single seed, a total of 1069 (865 as calibration set, 204 as validation set) single seeds representing 120 maize varieties were analyzed using chemical methods and performed calibration and external validation of the 150 single seeds set in parallel. Compared to various spectral pretreatments, the regression of partial least squares (PLS) with mathematical treatment of Harmonization showed the final optimization. The single-seed amylose contents showed the root mean square error of calibration (RMSEC) of 2.899, coefficient of determination for calibration (R2) of 0.902, and root mean square error of validation (RMSEV) of 2.948. In external validations, the coefficient of determination in cross-validation (r2), root mean square error of the prediction (RMSEP) and ratio of the standard deviation to SEP (RPD) were 0.892, 2.975 and 3.086 in the range of 20–30%, respectively. Therefore, NIRS will be helpful to breeders for determining the amylose contents of single-grain maize.

Accurate Quantification of Formaldehyde at ppb Level for Indoor Air Quality Monitoring

Accurate Quantification of Formaldehyde at ppb Level for Indoor Air Quality Monitoring