Build Prediction Models Research Articles

A Study on the Predictions of Wave Breaker Index in a Gravel Beach Using Linear Machine Learning Model

To date, numerous empirical formulas have been proposed through hydraulic model experiments to predict the wave breaker index, including wave height and depth of wave breaking, due to the inherent complexity of generation mechanisms. Unfortunately, research on the characteristics of wave breaking and the prediction of the wave breaker index for gravel beaches has been limited. This study aims to forecast the wave breaker index for gravel beaches using representative linear-based machine learning techniques known for their high predictive performance in regression or classification problems across various research fields. Initially, the applicability of existing empirical formulas for wave breaker indices to gravel seabeds was assessed. Various linear-based machine learning algorithms were then employed to build prediction models, aiming to overcome the limitations of existing empirical formulas in predicting wave breaker indices for gravel seabeds. Among the developed machine learning models, a new calculation formula for easily computable wave breaker indices based on the model was proposed, demonstrating high predictive performance for wave height and depth of wave breaking on gravel beaches. The study validated the predictive capabilities of the proposed wave breaker indices through hydraulic model experiments and compared them with existing empirical formulas. Despite its simplicity as a polynomial, the newly proposed empirical formula for wave breaking indices in this study exhibited exceptional predictive performance for gravel beaches.

Predicting E-Commerce CLV with Neural Networks: The Role of NPS, ATV, and CES

Accurately predicting Customer Lifetime Value (CLV) is paramount in optimizing customer relationship management. This study introduces a novel deep learning approach, employing a neural network model to forecast CLV for a leading international e-commerce retailer. This investigation delves into the interplay of key performance indicators—Net Promoter Score (NPS), Average Transaction Value (ATV), and Customer Effort Score (CES)—and their collective impact on CLV.The research utilized a dataset of 15,000 customer profiles and crafted a sequential neural network with dense layers optimized through hyperparameter tuning and regularization to thwart overfitting. The model's efficacy was assessed on a 10% test set, revealing its adeptness at capturing intricate, nonlinear predictor-customer lifetime value (CLV) relationships. Consistency in training and test performance metrics underscored the model's generalizability, while high R-squared and explained variance scores confirmed the predictive strength of the chosen factors.This research seeks to provide a solution for building an artificial intelligence model with an artificial neural network algorithm to predict customer lifetime value. By incorporating this principle, the study aims to leverage the power of neural networks to forecast CLV, enabling retailers to make informed decisions and optimize customer relationships accurately.This study finds that Net Promoter Score (NPS) and Customer Effort Score (CES) have a powerful impact on the neural network model's ability to predict Customer Lifetime Value (CLV) accurately. On the other hand, while Average Transaction Value (ATV) exhibits the least impact, it still significantly contributes to the accuracy of Customer Lifetime Value (CLV) predictions. These results underscore the importance of incorporating customer feedback metrics like NPS and CES when building predictive models for CLV estimation. Integrating critical customer metrics into neural networks gives retailers enhanced insights, enabling precise customer segmentation, resource allocation, and strategic growth. The study paves the way for future research to refine further CLV prediction, including dataset expansion, customer profile enrichment, and prescriptive marketing optimization.

Построение краткосрочного прогноза количества вагонов на станции и путях необщего пользования

Objective: collect raw data for building predictive models. Analyze the initial data, identify data outliers and outliers, divide the data into time intervals, calculate correlation coefficients, partial autocorrelation, cross-correlation, analyze the trend and seasonality of the obtained time series. Using autoregressive models, machine learning models, neuro-fuzzy models to build forecasts of time series and determine the quality of the obtained forecasts. Methods: point density, autocorrelation, partial autocorrelation, cross-correlation, Foster-Stewart test, Dickey-Fuller test, ARMA, MLP, Encoder-Decoder LTSM, TSK, Fuzzy-Partitions, SCRG, Transformers. Results: we obtained estimates of the prediction accuracy of the selected models, compared the results of the predictive models trained on different samples of initial data. Conclusions are made about the efficiency and methods of building predictive models. Practical significance: the 
 significance of building accurate predictive models for the key quantitative indicators of stations and nonpublic routes operation is shown. The factors influencing the accuracy of the obtained forecasts are analyzed.

Prediction of pathological complete response in locally advanced head and neck squamous cell carcinoma treated with neoadjuvant chemo-immunotherapy using volumetric multisequence MRI histogram analysis.

This study aimed to develop a multisequence MRI-based volumetric histogram metrics model for predicting pathological complete response (pCR) in advanced head and neck squamous cell carcinoma (HNSCC) patients undergoing neoadjuvant chemo-immunotherapy (NCIT) and compare its predictive performance with AJCC staging and RECIST 1.1 criteria. Twenty-four patients with locally advanced HNSCC from a prospective phase II trial were enrolled for analysis. All patients underwent pre- and post-NCIT MRI examinations from which whole-tumor histogram features were extracted, including T1WI, T2WI, enhanced T1WI (T1Gd), diffusion-weighted imaging (DWI) sequences, and their corresponding apparent diffusion coefficient (ADC) maps. The pathological results divided the patients into pathological complete response (pCR) and non-pCR (N-pCR) groups. Delta features were calculated as the percentage change in histogram features from pre- to post-treatment. After data reduction and feature selection, logistic regression was used to build prediction models. ROC analysis was performed to assess the diagnostic performance. Eleven of 24 patients achieved pCR. Pre_T2_original_firstorder_Minimum, Post_ADC_original_firstorder_MeanAbsoluteDeviation, and Delta_T1Gd_original_firstorder_Skewness were associated with achieving pCR after NCIT. The Combined_Model demonstrated the best predictive performance (AUC 0.95), outperforming AJCC staging (AUC 0.52) and RECIST 1.1 (AUC 0.72). The Pre_Model (AUC 0.83) or Post-Model (AUC 0.83) had a better predictive ability than AJCC staging. Multisequence MRI-based volumetric histogram analysis can non-invasively predict the pCR status of HNSCC patients undergoing NCIT. The use of histogram features extracted from pre- and post-treatment MRI exhibits promising predictive performance and offers a novel quantitative assessment method for evaluating pCR in HNSCC patients receiving NCIT.

A novel adaptive deep transfer learning method towards thermal error modeling of electric spindles under variable conditions

Thermal error modeling (TEM) plays a vital role in maintaining the machining accuracy of electric spindles. Recently, deep learning (DL) techniques have obtained promising achievements in this area. However, DL techniques have certain limitations. The data acquired from variable working conditions present large distribution discrepancies, the DL-based model established on one working condition fails to obtain satisfactory prediction accuracy in another working condition. Moreover, existing studies focus solely on using temperature features to build prediction models, neglecting the full use of multi-sensory information. To address these issues, this paper proposes a novel adaptive deep transfer learning method towards TEM of electric spindles, which takes full advantage of the temperature, current, and power sensory information. Firstly, finite element simulation is employed to analyze the thermal characteristics of the electric spindle and determine the locations for temperature measurement points. Then, the convolutional long short-term memory network (C-LSTMN) is constructed, where the spatial features from multi-sensory information between inputs and prediction patterns are captured by convolutional layers, these features are further processed by long short-term memory network (LSTMN) to extract temporal features. Subsequently, the multi-kernel joint maximum mean discrepancy (MK-JMMD) measure is developed to minimize the distribution discrepancies between the source and target domains, thus the prediction model initially established on the source domain can be adaptive to effectively predict on the target domain. Finally, with unavailable thermal error data in the target domain, the proposed method is validated through 12 transfer tasks using datasets from four working conditions with other comparison methods. The results demonstrate that the proposed method overcomes the challenges of unavailable labeled thermal error samples and outperforms advanced methods.

APRENDIZAJE AUTOMÁTICO PARA EL DIAGNÓSTICO MÉDICO PREDICTIVO, APLICACIÓN EN ZONAS INDUSTRIALES

Environmental pollution is a risk factor for chronic diseases (CD), which today are identified as the main cause of death in the world, 80% in low- and middle-income countries, in people of all ages. Given that industrial areas maintain a high rate of air pollution, their inhabitants are considered highly vulnerable, such is the case of the Metropolitan Area of Tula Hgo., in Mexico. At the request of the World Health Organization to integrate vulnerable populations to the quality of life through innovative strategies, the present study aims to build prediction models for CD of higher frequency in the area, to support predictive diagnosis through machine learning algorithms, recognized for their high performance in health areas. Based on the CRISP-DM methodology, requirements, characteristics and data behavior were analyzed, exhaustive cleaning and minmax scaler normalization were performed, the models were trained and validated with 80% - 20% of records, dropout and early stopping were applied to combat overtraining. The comparative analysis between 9 built models demonstrated the best performance of 3 of them, one for each EC; the Artificial Neural Network (ANN) for respiratory diseases and Random Forest (RF) for diabetes and high blood pressure. Its results of accuracy, precision, sensitivity, specificity and F1-score were 99%, 99%, 100%, 99% and 99.49% respectively for ANN, the RF model for diabetes obtained 98%, 100%, 97%, 100% and 98.7% and for arterial hypertension 95%, 97%, 94, 97% and 95.47%, these models were integrated into a graphical interface. The proposal constitutes a high-precision technological strategy for prevention and early diagnosis of CD in industrial areas, aimed at reducing mortality and improving the quality of life of the inhabitants.

Development of a nomogram for predicting acute pain among patients after abdominal surgery: A prospective observational study.

To develop a nomogram to provide a screening tool for recognising patients at risk of post-operative pain undergoing abdominal operations. Risk prediction models for acute post-operative pain can allow initiating prevention strategies, which are valuable for post-operative pain management and recovery. Despite the increasing number of studies on risk factors, there were inconsistent findings across different studies. In addition, few studies have comprehensively explored predictors of post-operative acute pain and built prediction models. A prospective observational study. A total of 352 patients undergoing abdominal operations from June 2022 to December 2022 participated in this investigation. A nomogram was developed for predicting the probability of acute pain after abdominal surgery according to the results of binary logistic regression. The nomogram's predictive performance was assessed by discrimination and calibration. Internal validation was performed via Bootstrap with 1000 re-samplings. A total of 139 patients experienced acute post-operative pain following abdominal surgery, with an incidence of 39.49%. Age <60, marital status (unmarried, divorced, or widowed), consumption of intraoperative remifentanil >2 mg, indwelling of drainage tubes, poor quality sleep, high pain catastrophizing, low pain self-efficacy, and PCIA not used were predictors of inadequate pain control in patients after abdominal surgery. Using these variables, we developed a nomogram model. All tested indicators showed that the model has reliable discrimination and calibration. This study established an online dynamic predictive model that can offer an individualised risk assessment of acute pain after abdominal surgery. Our model had good differentiation and calibration and was verified internally as a useful tool for risk assessment. The constructed nomogram model could be a practical tool for predicting the risk of experiencing acute post-operative pain in patients undergoing abdominal operations, which would be helpful to realise personalised management and prevention strategies for post-operative pain. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines were adopted in this study. Before the surgery, research group members visited the patients who met the inclusion criteria and explained the purpose and scope of the study to them. After informed consent, they completed the questionnaire. The patients' pain scores (VAS) were regularly assessed and documented by the bedside nurse for the first 3 days following surgery. Other information was obtained from medical records.

Theoretical-Methodological Foundations for the Global Integration Method (Método de Integração Global-MIG) in the Treatment of Autism Spectrum Disorder.

Currently, there is no intervention model for autism spectrum disorder (ASD) that addresses all levels and factors of the International Classification of Functioning, Disability and Health (ICF, WHO). The most researched programs focus on naturalistic, developmental and behavioral approaches to socio-communication. Less attention has been paid to motor and environmental reactivity aspects (behavior/interest restriction and sensory reactivity). The evidence rationale for the Global Integration Method (MIG, "Método de Integração Global"), a model addressing sensorimotor reactivity in addition to socio-communication, is presented. MIG is an integrative, interdisciplinary, family-oriented intervention and naturalistic program that addresses all levels and moderating factors of ASD's impact. MIG's theoretical rationale is based on the predictive coding impairment and embodied cognition hypotheses. MIG incorporates both bottom-up (flexible therapeutic suit, social-motor synchronization) and top-down (schematic social information processing, narratives, imagery) strategies to promote the building and use of accurate, flexible and context-sensitive internal predictive models. MIG is based on the premises that predictive coding improves both socio-communication and environmental reactivity, and that the postural stabilization provided by the flexible therapeutic suit frees information processing resources for socio-cognitive learning. MIG builds on interdisciplinary, professionally and parentally mediated work based on behavioral principles of intensive training in a situated environment.

Geographical distribution of two major quarantine fruit flies (Bactrocera minax Enderlein and Bactrocera dorsalis Hendel) in Sichuan Basin based on four SDMs.

Both Bactrocera minax and Bactrocera dorsalis are phytophagous insects, and their larvae are latent feeders, which cause great damage and economic losses to agriculture production and trade. This study aimed to provide a scientific reference for researching and developing the feasible countermeasures against these two pests. Based on the distribution data of B. minax and B. dorsalis in China, obtained from the Chinese herbaria, investigation and literature. Four niche models (Garp, Bioclim, Domain, and Maxent) were used to analyze the key environmental factors affecting the distribution of both pests and to build prediction models of the potential distribution in Sichuan Basin. Combined with two statistical standards, area under the receiver operating characteristic curve (AUC) and Kappa, the validity of prediction models were analyzed and compared. The results show that: the average AUC values of the four models are all above 0.90, and the average Kappa values are all above 0.75, indicating that the four models are suitable for predicting the potential distribution area of B. minax and B.dorsalis. The annual range of temperature, the mean temperature in the driest quarter, the mean temperature in the warmest quarter, the annual precipitation, and the precipitation in driest month are the key environmental factors affecting the distribution of B. minax, while the mean diurnal temperature range, the mean temperature in the driest quarter, the seasonal temperature variations and the precipitation in driest month affect the potential distribution of B. dorsalis. The suitable areas for B. minax are mainly concentrated in the eastern of Sichuan Basin, while the suitable areas for B. dorsalis are concentrated in the southeastern. Except for the Bioclim model, the highly-suitable area for both pests predicted by the other three models are all greater than 15.94 × 104 km2 and the moderately-suitable areas are greater than 13.57 × 104 km2. In conclusion, the suitable areas for both pests in Sichuan Basin are quite wide. Therefore, the relevant authorities should be given strengthened monitoring of both pests, especially in areas with high incursion rates.

Extended-Window Algorithms for Model Prediction Applied to Hybrid Power Systems

This paper proposes extended-window algorithms for model prediction and applies them to optimize hybrid power systems. We consider a hybrid power system comprising solar panels, batteries, a fuel cell, and a chemical hydrogen generation system. The proposed algorithms enable the periodic updating of prediction models and corresponding changes in system parts and power management based on the accumulated data. We first develop a hybrid power model to evaluate system responses under different conditions. We then build prediction models using five artificial intelligence algorithms. Among them, the light gradient boosting machine and extreme gradient boosting methods achieve the highest accuracies for predicting solar radiation and load responses, respectively. Therefore, we apply these two models to forecast solar and load responses. Third, we introduce extended-window algorithms and investigate the effects of window sizes and replacement costs on system performance. The results show that the optimal window size is one week, and the system cost is 13.57% lower than the cost of the system that does not use the extended-window algorithms. The proposed method also tends to make fewer component replacements when the replacement cost increases. Finally, we design experiments to demonstrate the feasibility and effectiveness of systems using extended-window model prediction.

Improved quantitative analysis of tenuifolin using hydrolytic continuous-flow system to build prediction models for its content based on near-infrared spectroscopy.

This study used two types of analyses and statistical calculations on powdered samples of Polygala root (PR) and Senega root (SR): (1) determination of saponin content by an independently developed quantitative analysis of tenuifolin content using a flow reactor, and (2) near-infrared spectroscopy (NIR) using crude drug powders as direct samples for metabolic profiling. Furthermore, a prediction model for tenuifolin content was developed and validated using multivariate analysis based on the results of (1) and (2). The goal of this study was to develop a rapid analytical method utilizing the saponin content and explore the possibility of quality control through a wide-area survey of crude drugs using NIR spectroscopy. Consequently, various parameters and appropriate wavelengths were examined in the regression analysis, and a model with a reasonable contribution rate and prediction accuracy was successfully developed. In this case, the wavenumber contributing to the model was consistent with that of tenuifolin, confirming that this model was based on saponin content. In this series of analyses, we have succeeded in developing a model that can quickly estimate saponin content without post-processing and have demonstrated a brief way to perform quality control of crude drugs in the clinical field and on the market.

Antimalarial Drug Combination Predictions Using the Machine Learning Synergy Predictor (MLSyPred©) tool

PurposeAntimalarial drug resistance is a global public health problem that leads to treatment failure. Synergistic drug combinations can improve treatment outcomes and delay the development of drug resistance. Here, we describe the implementation of a freely available computational tool, Machine Learning Synergy Predictor (MLSyPred©), to predict potential synergy in antimalarial drug combinations.MethodsThe MLSyPred© synergy prediction method extracts molecular fingerprints from the drugs’ biochemical structures to use as features and also cleans and prepares the raw data. Five machine learning algorithms (Logistic Regression, Random Forest, Support vector machine, Ada Boost, and Gradient Boost) were implemented to build prediction models. Implementation and application of the MLSyPred© tool were tested using datasets from 1540 combinations of 79 drugs and compounds biologically evaluated in pairs for three strains of Plasmodium falciparum (3D7, HB3, and Dd2).ResultsThe best prediction models were obtained using Logistic Regression for antimalarials with the strains Dd2 and HB3 (0.81 and 0.70 AUC, respectively) and Random Forest for antimalarials with 3D7 (0.69 AUC). The MLSyPred© tool yielded 45% precision for synergistically predicted antimalarial drug combinations that were annotated and biologically validated, thus confirming the functionality and applicability of the tool.Conclusion The MLSyPred© tool is freely available and represents a promising strategy for discovering potential synergistic drug combinations for further development as novel antimalarial therapies.

Efficient Use of Data for Prediction and Validation

Prediction model building is one of the most important tasks in analysis of high-dimensional data. A fitted prediction model should be validated for future use. So, when conducting such an analysis, we have to use the whole data for both training and validation. When using a hold-out method, the fitted prediction model will be more efficient if the training set is bigger, but the validation power will be lower with a smaller validation set. In order to balance the efficiency of fitted prediction model and its validation, 50-50 allocation of the whole data set is popularly used as a hold-out method. In prediction and validation procedure, we have to use the information embedded in the whole data set as efficiently as possible. As a such effort, cross-validation methods (CV) have been very popular these days. In a CV method, a large portion of the data set is used to train models and the remaining small portion of the data is used for validation, and this procedure is repeated until the whole data set is used for validation. In a CV method, each data point is used for both training and validation, so that as the portion of training set is increased, the efficiency of training will be increased, while the validation power will be decreased due to the increased over-fitting, i.e. more frequent use of each data point for training. As another effort of efficient use of the whole data, we propose to use the whole data set for both training and validation, called 1-fold CV method. By using the whole data to fit a prediction model, training efficiency will be highest, but, by reusing the whole data set for validation, its validation power is expected to be very low. The validation power of CV methods will be estimated by permutation methods. Through extensive simulation and real data studies, we conclude that the newly proposed 1-fold CV method uses the available data set very efficiently

Cancer Radiomic and Perfusion Imaging Automated Framework: Validation on Musculoskeletal Tumors.

Limitations from commercial software applications prevent the implementation of a robust and cost-efficient high-throughput cancer imaging radiomic feature extraction and perfusion analysis workflow. This study aimed to develop and validate a cancer research computational solution using open-source software for vendor- and sequence-neutral high-throughput image processing and feature extraction. The Cancer Radiomic and Perfusion Imaging (CARPI) automated framework is a Python-based software application that is vendor- and sequence-neutral. CARPI uses contour files generated using an application of the user's choice and performs automated radiomic feature extraction and perfusion analysis. This workflow solution was validated using two clinical data sets, one consisted of 40 pelvic chondrosarcomas and 42 sacral chordomas with a total of 82 patients, and a second data set consisted of 26 patients with undifferentiated pleomorphic sarcoma (UPS) imaged at multiple points during presurgical treatment. Three hundred sixteen volumetric contour files were processed using CARPI. The application automatically extracted 107 radiomic features from multiple magnetic resonance imaging sequences and seven semiquantitative perfusion parameters from time-intensity curves. Statistically significant differences (P < .00047) were found in 18 of 107 radiomic features in chordoma versus chondrosarcoma, including six first-order and 12 high-order features. In UPS postradiation, the apparent diffusion coefficient mean increased 41% in good responders (P = .0017), while firstorder_10Percentile (P = .0312) was statistically significant between good and partial/nonresponders. The CARPI processing of two clinical validation data sets confirmed the software application's ability to differentiate between different types of tumors and help predict patient response to treatment on the basis of radiomic features. Benchmark comparison with five similar open-source solutions demonstrated the advantages of CARPI in the automated perfusion feature extraction, relational database generation, and graphic report export features, although lacking a user-friendly graphical user interface and predictive model building.

Building and Validation of an Acute Event Prediction Model for Severe Mental Disorders.

The global incidence of acute events in psychiatric patients is intensifying, and models to successfully predict acute events have attracted much attention. To explore the influence factors of acute incident severe mental disorders (SMDs) and the application of Rstudio statistical software, and build and verify a nomogram prediction model. SMDs were taken as research objects. The questionnaire survey method was adopted to collect data. Patients with acute event independent factors were screened. R software multivariable Logistic regression model was constructed and a nomogram was drawn. A total of 342 patients with SMDs were hospitalized, and the number of patients who encountered acute events was 64, which accounted for 18.70% of all patients. Statistical significances were found in many aspects (all P ˂ 0.05). Such aspects included Medication adherence, disease diagnosis, marital status, caregivers, social support and the hospitalization environment (odds ratio (OR) = 4.08, 11.62, 12.06, 10.52, 0.04 and 0.61, respectively) were independent risk factors for the acute events of patients with SMDs. The prediction model was modeled, and the AUC was 0.77 and 0.80. The calibration curve shows that the model has good calibration. The clinical decision curve shows that the model has a good clinical effect. The constructed risk prediction model shows good prediction effectiveness in the acute events of patients with SMDs, which is helpful for the early detection of clinical mental health staff at high risk of acute events.

Viticultural Climate Indexes and Their Role in The Prediction of Anthocyanins and Other Flavonoids Content in Seedless Table Grapes

Background: Viticulture bioclimatic indexes like the Heliothermal Index (HI), Cool Night Index (CI), and Dryness Index (DI), can be used to assess the influence of climate on grapes’ quality. Methods: HI, CI, and DI + total seasonal irrigation were utilized to assess the effect of climate on the flavonoids content and composition of two Vitis vinifera seedless varieties, ‘Summer Royal’ and ‘Crimson Seedless’, both grown in Apulia (Southern Italy). Results: The flavonoids content was significantly affected by variety and climate conditions on the base of HI, CI, and DI + total seasonal irrigation. Factor analysis applied to climate indexes and flavonoids showed that anthocyanins and flavonols were negatively and positively correlated to CI in both varieties, respectively. Additionally, warmer night temperatures determined higher fla-van-3-ols. HI increase promoted anthocyanins, flavonols, and flavan-3-ols content in Crimson Seedless, whilst it induced an opposite trend in Summer Royal. Finally, DI + total seasonal irrigation showed to be positively linked to flavonols content and negatively linked to anthocyanins content just in the case of Crimson Seedless. Significant regression models were also determined between climate indexes and productive parameters (i.e., yield, TSS, TA, pH, bunch, and berry weight). Conclusions: Climate indexes HI, CI, and DI + total seasonal irrigation showed an effect on quality grape parameters like flavonoids and contributed to building predictive models when new climatic zones are going to be evaluated for the production of table grapes.

Prediction of soluble solid content in Nanfeng mandarin by combining hyperspectral imaging and effective wavelength selection

Soluble solid content (SSC) is among primary evaluation indicators of fruit quality and a key factor influencing consumer purchasing decisions. The research utilized hyperspectral imaging (380–1030 nm) to forecast SSC in Nanfeng mandarin. After a series of preprocessing methods, partial least squares regression (PLSR) and least squares support vector machine (LSSVM) were adopted to build prediction models. Combining multiplicative scatter correction and Savitzky-Golay smoothing was more effective compared to other preprocessing methods. Effective wavelengths (EWs) were selected by using bootstrapping soft shrinkage (BOSS), competitive adaptive reweighted sampling (CARS), iteratively retaining informative variables (IRIV) and their combinations. The BOSS-CARS-PLSR model performs optimally in prediction with the Rp2, RMSEP and RPD being 0.9376, 0.3986 and 4.0542, respectively. Additionally, the spatial distribution of the SSC in Nanfeng mandarin was visualized using the optimal model. Results show that combining hyperspectral imaging and EWs selection offers a rapid and intuitive approach that can non-destructively evaluate internal quality of Nanfeng mandarin.

Diagnostic value of adenohypophyseal MRI features in female children with precocious puberty

To evaluate the diagnostic value of adenohypophyseal magnetic resonance imaging (MRI) features for precocious puberty (PP) in female children and also to establish a non-invasive diagnostic approach in clinics. A total of 126 female children (37, 57, and 32 female children clinically diagnosed with central PP [CPP], incomplete PP [IPP], and controls, respectively) were enrolled in this study. Data were collected and analysed using analysis of variance. Pearson correlation and stepwise multivariate linear regression analysis were used to examine the association and build prediction models. Receiver operating characteristic (ROC) analysis was used to evaluate the diagnostic efficacy. The values of adenohypophysis volume (aPV), adenohypophysis height (aPH), and signal-intensity ratio (SIR), height, weight, and seven laboratory testing characteristics were correlated closely with the activation status of the hypothalamic-pituitary-gonad axis in the different groups (all p<0.05). Model 1 including aPV, weight, and aPH and Model 2 including SIR, aPV, and height were built to obtain predicted luteinising hormone (LH; R2=0.271) and LH/follicle stimulating hormone (FSH; R2=0.311). ROC analysis showed the predicted LH, predicted LH/FSH, and aPV were the top 3 best predictors in distinguishing CPP from controls (AUC=0.969, 0.949, and 0.938) while predicted LH/FSH was the best predictor in distinguishing CPP from IPP and controls (AUC=0.829 and 0.828). The adenohypophysis volume itself and the prediction models including main adenohypophyseal MRI features increased diagnostic efficiency for PP and offered a non-invasive and credible diagnostic method.

Interaction-Based Prediction for Dynamic Multiobjective Optimization

Dynamic multiobjective optimization poses great challenges to evolutionary algorithms due to the change of optimal solutions or Pareto front with time. Learning-based methods are popular to extract the changing pattern of optimal solutions for predicting new solutions. They tend to use all variables as features (i.e., inputs) to build prediction models. However, there are usually some irrelevant and redundant variables, which increase training difficulty and decrease prediction accuracy. This paper proposes a new interaction-based prediction method, which captures the correlation of variables with prediction targets and selects the most relevant variables to build prediction models using neural networks. In particular, the interaction between variables is detected to remove redundant variables. In addition, a correction procedure is developed to further improve predicted solutions according to the prediction error in past environments. The predicted solutions are used to update the population according to a specifically designed update strategy. Integrating the interaction-based prediction (IP) method into the framework of multiobjective evolutionary algorithm based on decomposition (MOEA/D), a new algorithm named IP-DMOEA is put forward. Experimental results on a typical dynamic multiobjective test suite demonstrate the better performance of the proposed IPDMOEA than state-of-the-art algorithms in terms of convergence speed and solution quality. The proposed IP-DMOEA is also successfully applied to the multirobot task scheduling problem.

PREDICTING CHRONIC DISEASES IN HIGHLY POLLUTED INDUSTRIAL ENVIRONMENTS USING MACHINE LEARNING

ABSTRACT: Environmental pollution is a risk factor for chronic diseases (CD), which today are identified as the main cause of death in the world, 80% in low- and middle-income countries, in people of all ages. Given that industrial areas maintain a high rate of air pollution, their inhabitants are considered highly vulnerable, such is the case of the Metropolitan Area of Tula Hgo., in Mexico. At the request of the World Health Organization to integrate vulnerable populations to the quality of life through innovative strategies, the present study aims to build prediction models for CD of higher frequency in the area, to support predictive diagnosis through machine learning algorithms, recognized for their high performance in health areas. Based on the CRISP-DM methodology, requirements, characteristics and data behavior were analyzed, exhaustive cleaning and minmax scaler normalization were performed, the models were trained and validated with 80% - 20% of records, dropout and early stopping were applied to combat overtraining. The comparative analysis between 9 built models demonstrated the best performance of 3 of them, one for each EC; the Artificial Neural Network (ANN) for respiratory diseases and Random Forest (RF) for diabetes and high blood pressure. Its results of accuracy, precision, sensitivity, specificity and F1-score were 99%, 99%, 100%, 99% and 99.49% respectively for ANN, the RF model for diabetes obtained 98%, 100%, 97%, 100% and 98.7% and for arterial hypertension 95%, 97%, 94, 97% and 95.47%, these models were integrated into a graphical interface. The proposal constitutes a high-precision technological strategy for prevention and early diagnosis of CD in industrial areas, aimed at reducing mortality and improving the quality of life of the inhabitants.