Comparison of statistical indices for the evaluation of crop models performance

This study presents a comparison of the usual statistical methods used for crop model assessment. A case study was conducted using a data set from observations of the total dry weight in diploid potato crop, and six simulated data sets derived from the observationsaimed to predict the measured data. Statistical indices such as the coefficient of determination, the root mean squared error, the relative root mean squared error, mean error, index of agreement, modified index of agreement, revised index of agreement, modeling efficiency, and revised modeling efficiency were compared. The results showed that the coefficient of determination is not a useful statistical index for model evaluation. The root mean squared error together with the relative root mean squared error offer an excellent notion of how deviated the simulations are in the same unit of the variable and percentage terms, and they leave no doubt when evaluating the quality of the simulations of a model.

Automatic Detection of Impervious Surfaces from Remotely Sensed Data Using Deep Learning

The large scale quantification of impervious surfaces provides valuable information for urban planning and socioeconomic development. Remote sensing and GIS techniques provide spatial and temporal information of land surfaces and are widely used for modeling impervious surfaces. Traditionally, these surfaces are predicted by computing statistical indices derived from different bands available in remotely sensed data, such as the Landsat and Sentinel series. More recently, researchers have explored classification and regression techniques to model impervious surfaces. However, these modeling efforts are limited due to lack of labeled data for training and evaluation. This in turn requires significant effort for manual labeling of data and visual interpretation of results. In this paper, we train deep learning neural networks using TensorFlow to predict impervious surfaces from Landsat 8 images. We used OpenStreetMap (OSM), a crowd-sourced map of the world with manually interpreted impervious surfaces such as roads and buildings, to programmatically generate large amounts of training and evaluation data, thus overcoming the need for manual labeling. We conducted extensive experimentation to compare the performance of different deep learning neural network architectures, optimization methods, and the set of features used to train the networks. The four model configurations labeled U-Net_SGD_Bands, U-Net_Adam_Bands, U-Net_Adam_Bands+SI, and VGG-19_Adam_Bands+SI resulted in a root mean squared error (RMSE) of 0.1582, 0.1358, 0.1375, and 0.1582 and an accuracy of 90.87%, 92.28%, 92.46%, and 90.11%, respectively, on the test set. The U-Net_Adam_Bands+SI Model, similar to the others mentioned above, is a deep learning neural network that combines Landsat 8 bands with statistical indices. This model performs the best among all four on statistical accuracy and produces qualitatively sharper and brighter predictions of impervious surfaces as compared to the other models.

Statistical evaluation of mineral water from Valcea County area through quality indices

Significant global consumption of mineral water is fueled by perceived therapeutic and medicinal qualities, cultural habits and taste. In Valcea County are several resorts with such mineral springs, which can have multiple benefits for human health. For this reason, it is important to investigate the level of their pollution with heavy metals. The aim of this study was to detect the level of heavy metals present in the studied mineral waters, to evaluate the analytical results using environmental statistical indices, and to compare the results with the legislation in force. Thus, mineral water samples were collected from three resorts of Valcea County (Baile Olanesti, Calimanesti-Caciulata and Baile Govora). The analyzed metals (Cd, Cr, Fe, Mn, Ni, Pb and Zn) were determined with the ICP-EOS technique and the obtained results were compared to enforce legislation. Statistical analyses were applied and two environmental statistical indices, namely the metal distribution coefficient (kd) and the total environmental risk index (IER) were evaluated. The calculated values for the total environmental risk index were below 0 (IER ≤ 0), which indicates that none of the studied water samples pose a risk for the environment. The low values of the distribution coefficient indicate a low ability to bind the metal in particles and therefore, insignificant toxicity. The distribution coefficient calculated for zinc (1.12 L/kg) and lead (0.68 L/kg) in Baile Olanesti indicates a high capacity of the metals to bind in particles, compared to other metals.

Assessment of Heavy Metal Contamination in Sediments of the Shitalakhya River, Bangladesh

Sediment samples collected from the river Shitalakhya, Bangladesh, were analyzed using atomic absorption spectroscopy (AAS) to investigate site-to-site (spatial) and seasonal (i.e., dry, premonsoon, post-monsoon) variation of Cr, Mn, Ni, Cu and Zn. The mean concentrations of Cr, Mn, Ni, Cu and Zn were 22.37 ± 6.09, 612.59 ± 160.08, 54.11 ± 11.21, 50.36 ± 9.40 and 103.62 ± 62.74 mg/kg in the dry, 31.58 ± 5.22, 569.71 ± 112.16, 58.35 ± 7.82, 49.93 ± 17.36 and 110.88 ± 95.83 mg/kg in the pre-monsoon and 18.09± 6.32, 567.02 ± 115.55, 50.89 ± 6.58, 39.75 ± 4.56 and 55.22 ± 11.33 mg/kg in the post-monsoon, respectively. Based on the metals’ concentrations, no considerable difference was observed among the three seasons, but the concentrations were slightly elevated in the dry and pre-monsoon compared to that in the post-monsoon with respect to site-to-site variation. Among the metals examined, concentrations of Ni and Cu were elevated because of the use of oxides of these heavy metals as catalysts in the ammonia plant. The following statistical indices i.e., Pearson correlation matrix, geo-accumulation index (Igeo), contamination factor (Cf), degree of contamination (Cd), pollution load index (PLI) and ecological risk potential (RI) factors were taken into account to assess the heavy metals contamination of the sediments. According to the values of the statistical indices for Cr, Mn, Ni, Cu and Zn, it is concluded that the study area was with low contamination while concentrations of Ni and Cu were higher than the Threshold Effect Level (TEL) and Toxicity Reference Value (TRV) values suggesting unsafe to use the sediments for vegetation and other uses.

Modelling of Thin-Layer Drying of Osmo-Pretreated Red Bell Pepper

The paper observes a thin-layer drying behaviour of red bell pepper. The red bell pepper (192 samples) was pretreated in osmotic solution of salt of concentrations 5–20% (w/w) at osmotic solution temperatures (30–60 °C) and osmotic process durations (30–120 min) and dried at 60 °C in a locally fabricated convective dryer after preformation of osmotic dehydration pretreatment process. Experimental moisture content values obtained from the drying process were converted to moisture ratios. Seven existing thin-layer drying model equations were used for model equation fitting. The predicted and experimental (observed) moisture ratios were analysed statistically. The statistical indices and rules used to judge and select the model equation that would best describe the process were the highest values of coefficient of determination (R 2); the lowest values of chi-square (χ2), root mean square error (RMSE), and sum of squares error (SSE). Results showed that the two-term exponential model equation best described the drying behaviour of osmo-pretreated red bell pepper. The ranges of statistical indices of selected two-term exponential model equation are: R 2 (0.9389–0.9751), χ2 (0.0642–0.1503), RMSE (0.2032–0.1668), and SSE (0.6424–1.5027).

The additive role of diffusion-weighted magnetic resonance imaging to axillary nodal status evaluation in cases of newly diagnosed breast cancer

Background Axillary nodal status is crucial for the management of cases with recently diagnosed breast cancer; usually addressed via axillary ultrasonography (US) along with tissue sampling in case of suspicion. Axillary nodal dissection and sentinel biopsy may be done, but are rather invasive, carrying a potential complication risk, which raises the need for non-invasive, reliable, pre-operative axillary imaging. We aimed at evaluating the performance of diffusion-weighted imaging (DWI) regarding preoperative axillary evaluation, using functional information derived from diffusion capacity differences between benign and malignant tissue. We included 77 axillary nodes from 77 patients (age range 20–78 years, mean 50 ± 12.6 SD) in our prospective study, presenting with variable clinical breast complaints, all scoring BIRADS 4/5 on sonomammography (SM). They underwent axillary evaluation by both US and DW-MRI where US classified nodes into benign, indeterminate, or malignant by evaluating nodal size, shape, cortical thickness, and hilar fat. Qualitative DWI classified them into either restricted or not and a cut-off apparent diffusion coefficient (ADC) value was calculated to differentiate benign and malignant nodal involvement. Results for each modality were correlated to those of final histopathology, which served as the standard of reference. Results The calculated sensitivity, specificity, accuracy, PPV, and NPV for US was 100%, 36.6%, 75.3%, 71.2%, and 100%, respectively. Statistical indices for qualitative DWI were 76.6%, 63.3%, 76.6%, 63.3%, and 71.4%, respectively (P value < 0.001). The calculated cut off value for ADC between infiltrated and non-infiltrated nodes was 0.95 × 10−3 mm2/s concluding statistical indices of 76.6%, 63.3%, 76.6%, 63.3%, and 71.4%, respectively (P value < 0.001). Conclusion Combining DW-MRI to conventional US improves diagnostic specificity and overall accuracy of preoperative axillary evaluation of patients with recently discovered breast cancer.

Performance Comparison of Non-Linear Median Filter Built on MLP-ANN and Conventional MLP-ANN: Using Improved Dataset Training in Micro-Cell Environment

This research work explores the Levenberg- Marquardt training algorithm used for Artificial Neural Network (ANN) optimization during training and the Bayesian Regularization algorithm for the enhanced generalized trained network in training a designed non-linear vector median filter built on Multi-Layer Perceptron (MLP) ANN called model-1 and a conventional MLP ANN called model-2. The model-1 employed in the design helps in dataset de-noising to ensure the removal of unwanted signals for the improved training dataset. An early stopping method in the ratio of 80:10:10 for training, testing, and validation to overcome the problem of over-fitting during network training was employed. First-order statistical indices, the standard deviation, root mean squared error, mean absolute error, and correlation coefficient were adopted for network training analysis and comparative analysis of the designed model-1 and model-2, respectively. Two locations, Line-of-sight (location-1) and non-Line-of-Sight (location-2), were considered where the dataset was captured. The training results from the two locations for the two models demonstrated improved prediction of signal power loss using model-1 in comparison to model-2. For instance, the correlation coefficient, which shows the strength of the predicted value to the measured values (closer to 1) establishing a strong connection, gives 0.990 and 0.995 using model-1 for location-1, training with Lavenberg-Marquardt and Bayesian Regularization algorithm, respectively and 0.965 and 0.980 for model-2 using the same algorithms. It is seen that the Bayesian regularization algorithm, which optimizes the network in accordance with the Levenberg- Marquardt algorithm, gave better prediction results. The same sequence of improved perditions using designed model-1 in comparison to model-2 were seen with training results in location-2 while also adopting other employed 1st order statistical indices.

Causal relations of health indices inferred statistically using the DirectLiNGAM algorithm from big data of Osaka prefecture health checkups

Causal relations among many statistical variables have been assessed using a Linear non-Gaussian Acyclic Model (LiNGAM). Using access to large amounts of health checkup data from Osaka prefecture obtained during the six fiscal years of years 2012–2017, we applied the DirectLiNGAM algorithm as a trial to extract causal relations among health indices for age groups and genders. Results show that LiNGAM yields interesting and reasonable results, suggesting causal relations and correlation among the statistical indices used for these analyses.

Simulation and optimization of control system operation and surface water allocation based on system dynamics modeling

In the present study, the performance of Dez Reservoir, its regulatory and downstream diversion dams are studied using the System Dynamics method (SD) in the Vensim software. In the first step, all the effective parameters with their relationship were identified and then the simulation was carried out considering the minimum and maximum reservoir volume and downstream demand over an eight-year period (2006–2013). After model validation, the statistical indices of correlation coefficient (R2), root mean square error (RMSE), and standard error (SE) for the reservoir volume were calculated as 0.99, 3.1690, and 0.18, respectively. The statistical indices for the regulatory dam were 0.98, 7.51, and 0.13, respectively. These statistical indices were 0.92, 4.71, and 0.37 for the diversion dam. The achieved results showed that the SD-simulated releases met downstream demands and the SD method has high accuracy in simulating the performance of the surface tank system. Moreover, for a better evaluation of the results of the SD model, the eight-year Dez dam releases were optimized using the particle swarm optimization (PSO) algorithm, and the required deficiency values were evaluated in both SD and PSO models. Results showed that SD provides more appropriate planning in meeting the downstream demands than PSO.