RESUMEN

BACKGROUND: Due to the sparse encoding character of the human visual cortex and the scarcity of paired training samples for {images, fMRIs}, voxel selection is an effective means of reconstructing perceived images from fMRI. However, the existing data-driven voxel selection methods have not achieved satisfactory results. NEW METHOD: Here, a novel deep reinforcement learning-guided sparse voxel (DRL-SV) decoding model is proposed to reconstruct perceived images from fMRI. We innovatively describe voxel selection as a Markov decision process (MDP), training agents to select voxels that are highly involved in specific visual encoding. RESULTS: Experimental results on two public datasets verify the effectiveness of the proposed DRL-SV, which can accurately select voxels highly involved in neural encoding, thereby improving the quality of visual image reconstruction. COMPARISON WITH EXISTING METHODS: We qualitatively and quantitatively compared our results with the state-of-the-art (SOTA) methods, getting better reconstruction results. We compared the proposed DRL-SV with traditional data-driven baseline methods, obtaining sparser voxel selection results, but better reconstruction performance. CONCLUSIONS: DRL-SV can accurately select voxels involved in visual encoding on few-shot, compared to data-driven voxel selection methods. The proposed decoding model provides a new avenue to improving the image reconstruction quality of the primary visual cortex.

RESUMEN

This study addresses the need for accurate structural data regarding the toxicity of fragrances in sanitizers and disinfectants. We compare the predictive and descriptive (model stability) potential of multiple linear regression (MLR) and partial least squares (PLS) models optimized through variable selection (VS). A novel hybrid chaotic neural network algorithm with competitive learning (CCLNNA)-PLS modeling strategy can offer specific optimization with satisfactory results, even for a limited dataset. While also exploring the preliminary comparative analysis, the goal is to introduce an adapted novel CCLNNA optimization strategy for VS, inspired by neural networks, along with exploring the influence of the percentage of significant descriptors in the optimization function to enhance the final model's capabilities. We analyzed an available dataset of 24 molecules, incorporating ADMET and PaDEL descriptors as predictor variables, to explore the relationship between the response/target variable (pLC50) and the meticulously optimized set of descriptors. The suitability of the selected PLS models (cross- and external-validated accuracy combined with percentage of significant descriptors at a level equal to or >80 %) underscores the importance of expanding the dataset to amplify the validation protocols, thus enhancing future model reliability and environmental impact.

Asunto(s)

RESUMEN

Soil erodibility (K) is an essential component in estimating soil loss indicating the soil's susceptibility to detach and transport. Data Computing and processing methods, such as artificial neural networks (ANNs) and multiple linear regression (MLR), have proven to be helpful in the development of predictive models for natural hazards. The present case study aims to assess the efficiency of MLR and ANN models to forecast soil erodibility in Peninsular Malaysia. A total of 103 samples were collected from various sites and K values were calculated using the Tew equation developed for Malaysian soil. From several extracted parameters, the outcomes of correlation and principal component analysis (PCA) revealed the influencing factors to be used in the development of ANN and MLR models. Based on the correlation and PCA results, two sets of influencing factors were employed to develop predictive models. Two MLR (MLR-1 and MLR-2) models and four neural networks (NN-1, NN-2, NN-3, and NN-4) optimized using Levenberg-Marquardt (LM) and scaled conjugate gradient (SCG) were developed and evaluated. The model performance validation was conducted using the coefficient of determination (R2), mean squared error (MSE), root mean squared error (RMSE), and Nash-Sutcliffe efficiency coefficient (NSE). The analysis showed that ANN models outperformed MLR models. The R2 values of 0.446 (MLR-1), 0.430 (MLR-2), 0.894 (NN-1), 0.855 (NN-2), 0.940 (NN-3), and 0.826 (NN-4); MSE values of 0.0000306 (MLR-1), 0.0000315 (MLR-2), 0.0000158 (NN-1), 0.0000261 (NN-2), 0.0000318 (NN-3), and 0.0000216 (NN-4) suggested the higher accuracy and lower modelling error of ANN models as compared with MLR. This study could provide an empirical basis and methodological support for K factor estimation in the region.

RESUMEN

Air pollution has emerged as a significant global concern, particularly in urban centers. This study aims to investigate the temporal distribution of air pollutants, including PM2.5, PM10, and O3, utilizing multiple linear regression modeling. Additionally, the research incorporates the calculation of the Air Quality Index (AQI) and Autoregressive Integrated Moving Average (ARIMA) time series modeling to predict the AQI for PM2.5 and PM10. The concentrations and AQI values for PM2.5 ranged from 0 to 93.6 µg/m3 and 0 to 171, respectively, surpassing the Word Health Organization's (WHO) acceptable threshold levels. Similarly, concentrations and AQI values for PM10 ranged from 0.1 to 149.27 µg/m3 and 2-98 µg/m3, respectively, also exceeding WHO standards. Particulate matter pollution exhibited notable peaks during summer and winter. Key meteorological factors, including dew point temperature, relative humidity, and rainfall, showed a significant negative association with all pollutants, while ambient temperature exhibited a significant positive correlation with particulate matter. Multiple linear regression models of particulate matter for winter season demonstrated the highest model performance, explaining most of the variation in particulate matter concentrations. The annual multiple linear regression model for PM2.5 exhibited the most robust performance, explaining 60% of the variation, while the models for PM10 and O3 explained 45% of the variation in their concentrations. Time series modeling projected an increasing trend in the AQI for particulate matter in 2022. The precise and accurate results of this study serve as a valuable reference for developing effective air pollution control strategies and raising awareness of AQI in Myanmar.

Asunto(s)

RESUMEN

Malaria continues to be a major barrier to socioeconomic development in Africa, where its death rate is over 90%. The predictive power of many machine learning models-such as multi-linear regression (MLR), artificial neural networks (ANN), adaptive neuro-fuzzy inference systems (ANFISs) and Random Forest classifier-is investigated in this study using data from 2207 patients. The dataset was reduced from the initial dataset of thirty-two criteria samples to fifteen. Assessment measures such as the root mean square error (RMSE), mean square error (MSE), coefficient of determination (R2), and adjusted correlation coefficient R were used. ANFIS, Random Forest, MLR, and ANN are among the models. After training, ANN outperforms ANFIS (97%), MLR (92%), and Random Forest (68%) with the greatest R (99%) and R2 (99%), respectively. The testing stage confirms the superiority of ANN. The paper also presents a statistical forecasting sheet with few errors and excellent accuracy for MLR models. When the models are assessed with Random Forest, the latter shows the least results, thus broadening the modeling techniques and offering significant insights into the prediction of malaria and healthcare decision making. The outcomes of using machine learning models for precise and efficient illness prediction add to an expanding body of knowledge, assisting healthcare systems in making better decisions and allocating resources more effectively.

RESUMEN

Gastrodin is one of the most important biologically active components of Gastrodia elata, which has many health benefits as a dietary and health food supplement. However, gastrodin measurement traditionally relies on laboratory and sophisticated instruments. This research was aimed at developing a rapid and non-destructive method based on Fourier transform near infrared (FT-NIR) to predict gastrodin content in fresh Gastrodia elata. Auto-ordered predictors selection (autoOPS) and successive projections algorithm (SPA) were applied to select the most informative variables related to gastrodin content. Based on that, partial least squares regression (PLSR) and multiple linear regression (MLR) models were compared. The autoOPS-SPA-MLR model showed the best prediction performances, with the determination coefficient of prediction (Rp2), ratio performance deviation (RPD) and range error ratio (RER) values of 0.9712, 5.83 and 27.65, respectively. Consequently, these results indicated that FT-NIRS technique combined with chemometrics could be an efficient tool to rapidly quantify gastrodin in Gastrodia elata and thus facilitate quality control of Gastrodia elata.

RESUMEN

Identifying groundwater contamination sources and supervising groundwater quality conditions are urgently needed to protect the groundwater resources of coastal areas like Contai of India, as communities here are heavily relying on groundwater which deteriorates progressively. So current research aims to address in detail about origins and influencing factors of groundwater contamination, status, and monitoring water quality by employing extremely useful leading technologies like principal component and factor analyses (PCA/FA), groundwater quality index (GWQI), and multiple linear regression (MLR) that helps to simplify complicated works instead of the conventional methods. Eight groundwater quality parameters were evaluated here, such as pH, TH (total hardness), Tur (turbidity), EC (electrical conductivity), TDS (total dissolved solids), Mn (manganese), Fe (iron), and Cl (chloride) for 38 sites. Three principal components with ~ 81% of the total variance were extracted from the PCA/FA analysis. The origin of maximum loadings of each factor is identified as a result of saline water, disintegration and leaching process, organic or else biogenic activities, and lithogenic or otherwise non-lithogenic links through percolating water. GWQI results show that ~ 87% of the samples fall into the good category and ~ 13% of the samples fall into the poor to very poor category. A model consisting of Tur, Fe, EC, Mn, TH, and Cl as independent parameters is more feasible and is proposed to predict GWQI obtained from MLR analysis. This MLR model also suggests that turbidity with the highest beta coefficient (0.820) is a key contributor relative to the entire groundwater class in this affected area. The findings relating to this research may support the designer and officials in monitoring and protecting coastal groundwater resources like selected areas.

Asunto(s)

RESUMEN

Fluoride pollution is a major issue worldwide, posing health risks such as dental and skeletal fluorosis. This study was conducted in fluoride enrichment hard rock regions of Vaniyambadi and Ambur talks in Tirupathur district, Tamil Nadu. Four hundred eighty groundwater samples were collected from May 2021 to April 2022 and grouped as summer, southwest monsoon (SWM), northeast monsoon (NEM), and winter. Maximum concentration of fluoride was found to be 4 mg/L in the summer season in Vellakuttai region of Vaniyambadi taluk. The study aims to investigate the hydrogeochemical process and mechanism influencing groundwater chemistry and it also provides the confirmation of exploratory data analysis in groundwater quality using structural equation modeling. The Piper and Gibbs diagrams illustrate the rock-water interaction and anthropogenic sources that contribute to the NaHCO3 and NaCl-type waters, respectively. Multivariate statistical analysis such as hierarchical cluster analysis (HCA), principal component analysis (PCA), multiple linear regression (MLR), and structural equation modeling (SEM) has been carried out to determine the groundwater quality. HCA manifests the nature and sources of groundwater, whereas PCA divides all the physicochemical parameters into two PC loadings, accounting for 97.46%, 99.46%, 99.18%, and 98.93% of cumulative % of variance during the summer, SWM, NEM, and winter seasons, respectively. PC1 has a higher loading factor to Cl, Ca, and Mg, whereas PC2 has a higher loading factor to Na, HCO3, SO4, and NO3. The results of the MLR model provide higher accuracy in detecting the contamination factors associated with the environment and natural rocks. SEM revealed the goodness-of-fit indices 0.993, 0.999, 1.000, and 0.999 in summer, SWM, NEM, and winter, respectively. Hence, this study provides insight view of variation of fluoride concentration in groundwater in different seasons and also mentions the factors that influence fluoride concentration in Vaniyambadi and Ambur taluk.

Asunto(s)

RESUMEN

A QbD-DM3 strategy was used to design ketoprofen (KTF) optimized liquid (L-SNEDDS) and solid self-nanoemulsifying drug delivery systems (S-SNEDDS). Principal component analysis was used to identify the optimized L-SNEDDS containing Capmul® MCM NF, 10 % w/w; Kolliphor® ELP, 60 % w/w; and propylene glycol, 30 % w/w. The S-SNEDDS was manufactured by spray-drying a feed dispersion prepared by dissolving the optimized KTF-loaded L-SNEDDS in an ethanol-Aerosil® 200 dispersion. A Box Behnken design was employed to evaluate the effect of drug concentration (DC), Aerosil® 200 concentration (AC) and feed rate (FR) on maximizing percent yield (PY) and loading efficiency (LE). The optimal levels of DC, AC, and FR were 19.9 % w/w, 30.0 % w/w, and 15.0 %, respectively. The optimized S-SNEDDS was amorphous, and its dissolution showed a 2.37-fold increase in drug release compared to KTF in 0.1 HCl. An optimized independent spray-dried S-SNEDDS verification batch showed that the predicted and observed PY and LE were 70.49 % and 92.49 %, and 70.02 % and 91.27 %, respectively. The optimized L-SNEDDS and S-SNEDDS also met their quality target product profile criteria for globule size <100 nm, polydispersity index < 0.400, emulsification time < 30 s, and KTF L-SNEDDS solubility 100-fold greater than its water solubility.

Asunto(s)

RESUMEN

Polycyclic aromatic hydrocarbons (PAHs) and their oxygen/nitrogen derivatives released into the atmosphere can alternate between a gas phase and a particulate phase, further affecting their environmental behavior and fate. The gas/particulate partition coefficient (KP) is generally used to characterize such partitioning equilibrium. In this study, the correlation between log KP of fifty PAH derivatives and their n-octanol/air partition coefficient (log KOA) was first analyzed, yielding a strong linear correlation (R2 = 0.801). Then, Gaussian 09 software was used to calculate quantum chemical descriptors of all chemicals at M062X/6-311+G (d,p) level. Both stepwise multiple linear regression (MLR) and support vector machine (SVM) methods were used to develop the quantitative structure-property relationship (QSPR) prediction models of log KP. They yield better statistical performance (R2 > 0.847, RMSE < 0.584) than the log KOA model. Simulation external validation and cross validation were further used to characterize the fitting performance, predictive ability, and robustness of the models. The mechanism analysis shows intermolecular dispersion interaction and hydrogen bonding as the main factors to dominate the distribution of PAH derivatives between the gas phase and particulate phase. The developed models can be used to predict log KP values of other PAH derivatives in the application domain, providing basic data for their ecological risk assessment.

Asunto(s)

RESUMEN

Modern industrialization has led to the creation of a wide range of organic chemicals, especially in the form of multicomponent mixtures, thus making the evaluation of environmental pollution more difficult by normal methods. In this paper, we attempt to use forward stepwise multiple linear regression (MLR) and nonlinear radial basis function neural networks (RBFNN) to establish quantitative structure-activity relationship models (QSARs) to predict the toxicity of 79 binary mixtures of aquatic organisms using different hypothetical descriptors. To search for the proper mixture descriptors, 11 mixture rules were performed and tested based on preliminary modeling results. The statistical parameters of the best derived MLR model were Ntrain = 62, R2 = 0.727, RMS = 0.494, F = 159.537, Q2LOO = 0.727, and Q2pred = 0.725 for the training set; and Ntest = 17, R2 = 0.721, RMS = 0.508, F = 38.773, and q2ext = 0.720 for the external test set. The RBFNN model gave the following statistical results: Ntrain = 62, R2 = 0.956, RMS = 0.199, F = 1279.919, Q2LOO = 0.955, and Q2pred = 0.855 for the training set; and Ntest = 17, R2 = 0.880, RMS = 0.367, F = 110.980, and q2ext = 0.853 for the external test set. The quality of the models was assessed by validating the relevant parameters, and the final results showed that the developed models are predictive and can be used for the toxicity prediction of binary mixtures within their applicability domain.

Asunto(s)

RESUMEN

Given the great importance of cobalt catalysts supported by benchmark bis(imino)pyridine in the (oligo)polymerization, a series of dibenzopyran-incorporated symmetrical 2,6-bis(imino) pyridyl cobalt complexes (Co1-Co5) are designed and prepared using a one-pot template approach. The structures of the resulting complexes are well characterized by a number of techniques. After activation with either methylaluminoxane (MAO) or modified MAO (MMAO), the complexes Co1-Co4 are highly active for ethylene polymerization with a maximum activity of up to 7.36 × 106 g (PE) mol-1 (Co) h-1 and produced highly linear polyethylene with narrow molecular weight distributions, while Co5 is completely inactive under the standard conditions. Particularly, complex Co3 affords polyethylene with high molecular weights of 85.02 and 79.85 kg mol-1 in the presence of MAO and MMAO, respectively. The 1H and 13C NMR spectroscopy revealed the existence of vinyl end groups in the resulting polyethylene, highlighting the predominant involvement of the ß-H elimination reaction in the chain-termination process. To investigate the mechanism underlying the variation of catalytic activities as a function of substituents, multiple linear regression (MLR) analysis was performed, showing the key role of open cone angle (θ) and effective net charge (Q) on catalytic activity.

RESUMEN

The principal goal of the brain-computer interface (BCI) is to translate brain signals into meaningful commands to control external devices or neuroprostheses to restore lost functions of patients with severe motor disabilities. The invasive recording of brain signals involves numerous health issues. Therefore, BCIs based on non-invasive recording modalities such as electroencephalography (EEG) are safer and more comfortable for the patients. The BCI requires reconstructing continuous movement parameters such as position or velocity for practical application of neuroprostheses. The BCI studies in continuous decoding have extensively relied on extracting features from the amplitude of brain signals, whereas the brain connectivity features have rarely been explored. This study aims to investigate the feasibility of using phase-based connectivity features in decoding continuous hand movements from EEG signals. To this end, the EEG data were collected from seven healthy subjects performing a 2D center-out hand movement task in four orthogonal directions. The phase-locking value (PLV) and magnitude-squared coherence (MSC) are exploited as connectivity features along with multiple linear regression (MLR) for decoding hand positions. A brute-force search approach is employed to find the best channel pairs for extracting features related to hand movements. The results reveal that the regression models based on PLV and MSC features achieve the average Pearson correlations of 0.43 ± 0.03 and 0.42 ± 0.06, respectively, between predicted and actual trajectories over all subjects. The delta and alpha band features have the most contribution in regression analysis. The results also demonstrate that both PLV and MSC decoding models lead to superior results on our data compared to two recently proposed feature extraction methods solely based on the amplitude or phase of recording signals (p < 0.05). This study verifies the ability of PLV and MSC features in the continuous decoding of hand movements with linear regression. Thus, our findings suggest that extracting features based on brain connectivity can improve the accuracy of trajectory decoder BCIs.

RESUMEN

Evaluating and predicting date fruit quality during cold storage is critical for ensuring a steady supply of high-quality fruits to meet market demands. The traditional destructive methods take time in the laboratory, and the results are based on one specific parameter being tested. Modern modeling techniques, such as Machine Learning (ML) algorithms, offer unique benefits in nondestructive methods for food safety detection and predicting quality attributes. In addition, the electrical properties of agricultural products provide crucial information about the interior structures of biological tissues and their physicochemical status. Therefore, this study aimed to use an alternative approach to predict physicochemical properties, i.e., the pH, total soluble solids (TSS), water activity (aw), and moisture content (MC) of date fruits (Tamar), during cold storage based on their electrical properties using Artificial Neural Networks (ANNs), which is the most popular ML technique. Ten date fruit cultivars were studied to collect data for the targeted parameters at different cold storage times (0, 2, 4, and 6 months) to train and test the ANNs models. The electrical properties of the date fruits were measured using a high-precision LCR (inductance, capacitance, and resistance) meter from 10 Hz to 100 kHz. The ANNs models were compared with a Multiple Linear Regression (MLR) at all testing frequencies of the electrical properties. The MLR models were less accurate than ANNs models in predicting fruit pH and had low performance and weak predictive ability for the TSS, aw, and MC at all testing frequencies. The optimal ANNs prediction model consisted of the input layer with 14 neurons, one hidden layer with 15 neurons, and the output layer with 4 neurons, which was determined depending on the measurements of the electrical properties at a 10 kHz testing frequency. This optimal ANNs model was able to predict the pH with R2 = 0.938 and RMSE = 0.121, TSS with R2 = 0.954 and RMSE = 2.946, aw with R2 = 0.876 and RMSE = 0.020, and MC with R2 = 0.855 and RMSE = 0.803 b by using the measured electrical properties. The developed ANNs model is a powerful tool for predicting fruit quality attributes after learning from the experimental measurement parameters. It can be suggested to efficiently predict the pH, total soluble solids, water activity, and moisture content of date fruits based on their electrical properties at 10 kHz.

RESUMEN

BACKGROUND: Coronavirus outbreak (SARS-CoV-2) has become a serious threat to human society all around the world. Due to the rapid rate of disease outbreaks and the severe shortages of medical resources, predicting COVID-19 disease severity continues to be a challenge for healthcare systems. Accurate prediction of severe patients plays a vital role in determining treatment priorities, effective management of medical facilities, and reducing the number of deaths. Various methods have been used in the literature to predict the severity prognosis of COVID-19 patients. Despite the different appearance of the methods, they all aim to achieve generalizable results by increasing the accuracy and reducing the errors of predictions. In other words, accuracy is considered the only effective factor in the generalizability of models. In addition to accuracy, reliability and consistency of results are other critical factors that must be considered to yield generalizable medical predictions. Since the role of reliability in medical decisions is significant, upgrading reliable medical data-driven models requires more attention. METHODS: This paper presents a new modeling technique to specify and maximize the reliability of results in predicting the severity prognosis of COVID-19 patients. We use the well-known classic regression as the basic model to implement our proposed procedure on it. To assess the performance of the proposed model, it has been applied to predict the severity prognosis of COVID-19 by using a dataset including clinical information of 46 COVID-19 patients. The dataset consists of two types of patients' outcomes including mild (discharge) and severe (ICU or death). To measure the efficiency of the proposed model, we compare the accuracy of the proposed model to the classic regression model. RESULTS: The proposed reliability-based regression model, by achieving 98.6% sensitivity, 88.2% specificity, and 93.10% accuracy, has better performance than classic accuracy-based regression model with 95.7% sensitivity, 85.5% specificity, and 90.3% accuracy. Also, graphical analysis of ROC curve showed AUC 0.93 (95% CI 0.88-0.98) and AUC 0.90 (95% CI 0.85-0.96) for classic regression models, respectively. CONCLUSIONS: Maximizing reliability in the medical forecasting models can lead to more generalizable and accurate results. The competitive results indicate that the proposed reliability-based regression model has higher performance in predicting the deterioration of COVID-19 patients compared to the classic accuracy-based regression model. The proposed framework can be used as a suitable alternative for the traditional regression method to improve the decision-making and triage processes of COVID-19 patients.

Asunto(s)

RESUMEN

The Mar Menor is a Mediterranean coastal saltwater lagoon (Murcia, Spain) that represents a unique ecosystem of vital importance for the area, from both an economic and ecological point of view. During the last decades, the intense agricultural activity has caused episodes of eutrophication due to the contribution of inorganic nutrients, especially nitrates. For this reason, it is important to control the quality of the water discharged into the Mar Menor lagoon, which can be performed through the measurement of dissolved oxygen (DO). Therefore, this article aimed to predict the DO in the water discharged into this lagoon through the El Albujón watercourse, for which two theoretical models consisting of a multiple linear regression (MLR) and a back-propagation neural network (RPROP) were developed. Data of temperature, pH, nitrates, chlorides, sulphates, electrical conductivity, phosphates and DO at the mouth of this watercourse, between January 2014 and January 2021, were used. A preliminary statistical study was performed to discard the variables with the lowest influence on DO. Finally, both theoretical models were compared by means of the coefficient of determination (R2), the root mean square errors (RMSE) and the mean absolute error (MAE), concluding that the neural network made a more accurate prediction of DO.

Asunto(s)

RESUMEN

Repurposed used cooking oil is a sustainable alternative to other feedstocks for biodiesel production offering enviro-economic benefits. Residual crude glycerol (RCG) from such biodiesel production plants is difficult to utilize due to presence of numerous toxic impurities with various inhibitory effects on biological fermentative reforming process. However, it is a new industrial feedstock for bio-based production of 1,3-propanediol. In this work, a new Clostridium butyricum strain L4 was isolated from biogas reactor leachate after rigorous adaption and 35 subcultures under increasing stress conditions and studied for green production of 1,3-propanediol (PDO) from RCG and further process development. Evaluation of fermentative reforming kinetics was performed and the optimal reaction conditions are pH 7.0, temperature 30 °C, 2 g yeast extract/L and 15 g ammonium sulphate/L. Glycerol-glucose co-fermentation (10:1) enhanced cell growth and thus, PDO output by 11.6 g/L. In comparison to batch fermentation (24.8 g PDO/L; 0.58 mol PDO/mol glycerol) there was 2.8-fold improvement with fed-batch process resulting in accumulation of 70.1 g PDO/L (Yield = 0.65 mol PDO/mol glycerol) using the studied biocatalyst in 150 h. In order to predict yields under different operational conditions a multiple linear regression model was developed (r2 = 0.783) with six independent variables (p < 0.05), where biomass (g/L) and temperature (oC) were forecasted as top contributors to PDO yield. Finally, this biocatalyst appears as a potential candidate for industrial use due to its non-pathogenic nature, ability to grow in wide pH and temperature conditions, tolerance to high substrate and product concentration, insignificant generation of by-products and Coenzyme B12 independent biotransformation. The study can add value to bio-utilization of RCG to produce green 1,3-propanediol.

Asunto(s)

RESUMEN

This study involves information about the concentrations of nutrients (proteins, phenolic compounds, free amino acids, minerals (Ca, P, and Iron), hardness) in milled rice processed with enzymes; xylanase and cellulase produced by Aspergillus awamori, MTCC 9166 and Trichoderma reese, MTCC164. Brown rice was processed with 60-100% enzyme (40 mL buffer -undiluted) for 30 to 150 min at 30 °C to 50 °C followed by polishing for 20-100 s at a safe moisture level. Multiple linear regression (MLR) and artificial neural network (ANN) models were used for process optimization of enzymes. The MLR correlation coefficient (R2) varied between 0.87-0.90, and the sum of square (SSE) was placed within 0.008-8.25. While the ANN R2 (correlation coefficient) varied between 0.97 and 0.9999(1), MSE changed from 0.005 to 6.13 representing that the ANN method has better execution across MLR. The optimized cellulase process parameters (87.2% concentration, 80.1 min process time, 33.95 °C temperature and 21.8 s milling time) and xylanase process parameters (85.7% enzyme crude, 77.1 min process time, 35 °C temperature and 20 s) facilitated the increase of Ca (70%), P (64%), Iron (17%), free amino acids (34%), phenolic compounds (78%) and protein (84%) and decreased hardness (20%) in milled rice. Scanning electron micrographs showed an increased rupture attributing to enzymes action on milled rice.

RESUMEN

BACKGROUND AND AIMS: In recent decades, modeling and forecasting have played a significant role in the diagnosis and treatment of different diseases. Various forecasting models have been developed to improve data-based decision-making processes in medical systems. Although these models differ in many aspects, they all originate from the assumption that more generalizable results are achieved by more accurate models. This means that accuracy is considered as the only prominent feature to evaluate the generalizability of forecasting models. On the other side, due to the changeable medical situations and even changeable models' results, making stable and reliable performance is necessary to adopt appropriate medical decisions. Hence, reliability and stability of models' performance is another effective factor on the model's generalizability that should be taken into consideration in developing medical forecasting models. METHODS: In this paper, a new reliability-based forecasting approach is developed to address this gap and achieve more consistent performance in making medical predictions. The proposed approach is implemented on the classic regression model which is a common accuracy-based statistical method in medical fields. To evaluate the effectiveness of the proposed model, it has been performed by using two medical benchmark datasets from UCI and obtained results are compared with the classic regression model. RESULTS: Empirical results show that the proposed model has outperformed the classic regression model in terms of error criteria such as MSE and MAE. So, the presented model can be utilized as an appropriate alternative for the traditional regression model in making effective medical decisions. CONCLUSIONS: Based on the obtained results, the proposed model can be an appropriate alternative for traditional multiple linear regression for modeling in real-world applications, especially when more generalization and/or more reliability is needed.

Asunto(s)

RESUMEN

Statistical water quality forecast models are useful tools to assist with beach management. In particular, multiple linear regression (MLR) models have been successfully developed for prediction of fecal indicator bacteria concentrations for beaches in river, lake, and marine environments. Nevertheless, an unresolved challenging issue is the reliable prediction of infrequent events of high bacterial concentrations to inform beach closure decisions to protect public health. The number of field data available for the infrequent events is typically an order of magnitude less than that for days when the water quality criterion is met-MLR models often perform poorly in predicting bacterial concentrations on days when the beaches should be closed. For beach management in Hong Kong, MLR models have been developed to predict beach water quality indices in terms of four gradings (BWQI-1 to 4) based on Escherichia coli (E. coli) concentrations. In this study, we propose an artificial intelligence (AI)-based binary classification (EasyEnsemble) model using class-imbalance learning to predict "very poor" occasions (BWQI-4)-when E. coli concentration exceeds 610 counts/100 mL. Models are developed for three marine beaches with different hydrographic and pollution characteristics using a 30 year data set spanning three periods with different water quality status. The model-data comparison over a wide range of conditions shows that the proposed method results in a significant improvement in the prediction of "very poor" water quality. The proposed class-imbalance method for predicting rare events has an F-score of 0.84, and it significantly outperforms MLR and classification tree (CT) models with corresponding F-scores of 0.39 and 0.69. A robust beach water quality forecast system can hence be developed using hybrid MLR-binary classification modeling.

Asunto(s)