Cross-validated Values Research Articles

K-nearest neighbour and K-fold cross-validation used in wind turbines for false alarm detection

Maintenance management is critical to ensure the efficiency and competitiveness of energy production in the wind energy industry. Current maintenance operations are planned according to the alarm distribution of the wind farm provided by the supervisory control and data acquisition system, and false alarms may lead to unnecessary operation. The detection of false alarms is critical to ensure proper wind turbine maintenance management plans because false alarms increase the complexity and the costs of maintenance operations. This paper proposes a methodology based on K-nearest neighbour algorithms that compare different k-fold cross-validation values for false alarm detection. A real case scenario formed by three real wind turbines is presented to test the reliability of the methodology. The signal and alarm dataset are acquired by the supervisory control and data acquisition system and an alarm log as a response variable. It is demonstrated that the performance of the three wind turbines is analogous and the variations of the k-cross validation value showed that the accuracy does not increase significantly. The proposed methodology presented an accuracy of 98 % and more than 22 % of false alarms were detected in the case study. These outcomes demonstrate the robustness of the proposed approach for detecting false alarms in wind turbines.

Comparison of Raman and mid-infrared spectroscopy for quantification of nitric acid in PUREX-relevant mixtures

During the plutonium uranium reduction extraction (PUREX) process, nitric acid facilitates the extraction of actinides from the aqueous phase into the organic phase by forming neutral, organic soluble complexes with tri-n-butyl phosphate (TBP). The concentration of nitric acid is generally measured by titration; however, titration is a time-consuming method that generates significant volumes of additional waste. Optical spectroscopic techniques can be used to perform fast, automated measurements off-line or on-line, without generating any waste. In this work, the effectiveness of Raman and mid-infrared (MIR) spectroscopy has been compared for the first time as an alternative to titration for the quantification of nitric acid in PUREX-relevant mixtures. Samples of 0–12 M nitric acid in the aqueous phase and 0–1.10 M nitric acid in the organic phase (TBP/odourless kerosene (OK)–H2O–HNO3 model system) were analysed and partial least squares (PLS) regression models were built to predict nitric acid concentration. MIR spectra required less pre-processing than Raman spectra and more accurate predictions of nitric acid concentration were obtained for MIR spectroscopy than for Raman spectroscopy, with root mean square error of prediction (RMSEP) values of 0.099 M versus 0.148 M obtained for the aqueous phase and root mean square error of cross validation (RMSECV) values of 0.006 M versus 0.013 M obtained for the organic phase. To investigate the ability to predict nitric acid concentration in the presence of uranyl nitrate, samples containing uranium (0–100 g/L) and nitric acid (0.15–0.64 M) in the organic phase (U-TBP/OK-H2O–HNO3 model system) were analysed by Raman and MIR spectroscopy. The RMSECV was 0.027 M and 0.066 M for MIR and Raman spectroscopy, respectively; these values are higher than those obtained in the absence of uranyl nitrate owing to differences in the experimental approaches employed. Therefore, the results obtained demonstrate that MIR or Raman spectroscopy could be used to measure the concentration of nitric acid in the organic and aqueous phases in the PUREX process.

Weight estimation models for commercial Pinus radiata wood in small felling stands based on UAV-LiDAR data

Uncrewed Aerial Vehicles (UAVs) equipped with Light Detection and Ranging (LiDAR) scanners have been adopted as an effective tool for conducting forest inventories, reducing field work and associated costs. These studies are based on obtaining information on volume, height and diameter at breast height (DBH). However, in the timber sector in Galicia (Spain)the variable used as the unit of purchase/sale is "weight". Therefore, this paper sets out to test different machine learning algorithms, such as multiple linear regression (MLR), MLR log-transformed (MRL-LT), Principal Component Analysis (PCA) and Random Forest (RF) to obtain a methodology applicable in the study area and replicable in other areas, with real utility in the forestry industry for determining the weight of Pinus radiata wood. These models are based on 73 observations and 19 predictors and have been developed and applied to a total of 33 Forest Cuts (FC) divided into 5204 tesserae. The models are validated by Fold Cross-Validation, with RMSE (relative root mean square error), R2 and MAE (mean absolute error) values being calculated. Of the four models studied, the PCA model performs best (R2=0.85, RMSE=1.84, MAE=1.46), followed by the MRL-LT which gives a value of R2=0.82 (RMSE=1.84, MAE=1.46). The actual figure for total Pinus radiata timber harvested was 35,440.54 t The PCA model estimated a total of 35,913.53 t (r = 0.98, R2=0.97, εr=9.35 %), while the MRL-LT model calculated 35,847.92 t (r = 0.99, R2=0.98, εr=8.85 %). On the other hand, the weight was underestimated by the MLR model and overestimated by the RF. The plain MLR model underestimated the "weight" of the wood, while the MRL-LT model provided the best result, with errors of less than 10 % in 72 % of the FCs. In conclusion, this study provides a powerful tool that will enable stakeholders in the community timber industry to accurately estimate the weight of timber in Pinus radiata stands, enhancing a more highly automated forest inventory that can optimise the number of field operations to be performed.

SELECTION OF THE BEST B-SPLINE REGRESSION MODEL FOR ESTIMATING BITCOIN PRICE INCREASES BASED ON ORDER AND OPTIMAL KNOT POINT

The cryptographic virtual currency, bitcoin, is considered the main originator of cryptocurrencies that emerged due United States financial crisis in 2008. The idea was sparked by Nakamoto by introducing an alternative currency system that really refers to the strength of supply and demand. Based on INDODAX data, the bitcoin exchange rate during October 2020 to February 2021 is a condition of a large increase in a short time with a percentage increase of 450%. The increase in bitcoin prices can be modelled using the b-spline nonparametric regression method based on order and optimal knot points based on the smallest Generalized Cross Validation value. The resulting b-spline 4 degree and the number of knots points 5 as the best model with each bases described recursively.

Towards the automation of NIR spectroscopy to assess vineyard water status spatial–temporal variability from a ground moving vehicle

Irrigation has a strong impact in terms of yield regulation and grape and wine quality, so the implementation of precision watering systems would facilitate the decision-making process about the water use efficiency and the irrigation scheduling in viticulture. The objectives of this work were two-fold. On one hand, to compare and assess grapevine water status using two different spectral devices assembled in a mobile platform and to evaluate their capability to map the spatial variability of the plant water status in two commercial vineyards from July to early October in season 2021, and secondly to develop an algorithm capable of automate the spectral acquisition process using one of the two spectral sensors previously tested. Contemporarily to the spectral measurements collected from the ground vehicle at solar noon, stem water potential (Ψs) was used as the reference method to evaluate the grapevine water status. Calibration and prediction models for grapevine water status assessment were performed using the Partial least squares (PLS) regression and the Variable Importance in the Projection (VIP) method. The best regression models returned a determination coefficient for cross validation (R2cv) and external validation (R2p) of 0.70 and 0.75 respectively, and the standard error of cross validation (RMSECV) values were lower than 0.105 MPa and 0.128 MPa for Tempranillo and Graciano varieties using a more expensive and heavier near-infrared (NIR) spectrometer (spectral range 1200–2100 nm). Remarkable models were also built with the miniaturized, low-cost spectral sensor (operating between 900–1860 nm) ranging from 0.69 to 0.71 for R2cv, around 0.74 in both varieties for R2p and the RMSECV values were below 0.157 MPa, while the RMSEP values did not exceed 0.151 MPa in both commercial vineyards. This work also includes the development of a software which automates data acquisition and allows faster (up to 40% of time saving in the field) and more efficient deployment of the developed algorithm. The encouraging results presented in this work demonstrate the great potential of this methodology to assess the water status of the vineyard and estimate its spatial variability in different commercial vineyards, providing useful information for better irrigation scheduling.

CO and NOx emissions prediction in gas turbine using a novel modeling pipeline based on the combination of deep forest regressor and feature engineering

The main objective of this study is to estimate carbon oxides (CO) and nitrogen oxides (NOx) emissions from a gas turbine using the predictive emission monitoring systems dataset. First, four methods were developed for feature generation: Principal Component Analysis, t-Distributed Stochastic Neighbor Embedding, Uniform Manifold Approximation and Projection, and Potential of Heat-diffusion for Affinity-based Trajectory Embedding. Then Feature Relevance-based Unsupervised Feature Selection was evaluated for ranking features. With all features generated, the regression models Ridge Regression, Least Absolute Shrinkage and Selection Operator, k-Nearest Neighbor, Cubist Regression, Random Forest, Light Gradient Boosting Machine, Categorical Boosting, and Deep Forest Regression (DFR) were evaluated. The hyperparameters were tuned with Randomized Search Cross Validation using a 5-fold cross-validation (CV) procedure. In this paper, an innovative modeling pipeline based on the mentioned regressors, feature engineering, and effective hyperparameters tuning was proposed. The developed regression models based on feature engineering and hyperparameters tuning were compared using performance criteria, including mean 5-fold CV values of the coefficient of determination (R2) and root mean square error (RMSE). The obtained best results based on mean CV values using the DFR presented R2 values equal to 0.9647 and 0.5355 and RMSE values equal to 0.4245 and 1.4508 for CO emissions, for the training and validation datasets, respectively. On the other hand, DFR presented R2 equal to 0.9866 and 0.4737 and RMSE equal to 1.3876 and 7.6528 for NOx emissions, for the training and validation datasets, respectively.The results show that the proposed DFR model has higher potential for CO and NOx emissions prediction and that feature engineering and hyperparameter tuning has demonstrated impacts on the predictive capacity of the evaluated regression models.

Radiomics under 2D regions, 3D regions, and peritumoral regions reveal tumor heterogeneity in non-small cell lung cancer: a multicenter study.

Lung cancer has significant genetic and phenotypic heterogeneity, leading to poor prognosis. Radiomic features have emerged as promising predictors of the tumor phenotype. However, the role of underlying information surrounding the cancer remains unclear. We conducted a retrospective study of 508 patients with NSCLC from three institutions. Radiomics models were built using features from six tumor regions and seven classifiers to predict three prognostically significant tumor phenotypes. The models were evaluated and interpreted by the mean area under the receiver operating characteristic curve (AUC) under nested cross-validation and Shapley values. The best-performing predictive models corresponding to six tumor regions and three tumor phenotypes were identified for further comparative analysis. In addition, we designed five experiments with different voxel spacing to assess the sensitivity of the experimental results to the spatial resolution of the voxels. Our results demonstrated that models based on 2D, 3D, and peritumoral region features yielded mean AUCs and 95% confidence intervals of 0.759 and [0.747-0.771] for lymphovascular invasion, 0.889 and [0.882-0.896] for pleural invasion, and 0.839 and [0.829-0.849] for T-staging in the testing cohort, which was significantly higher than all other models. Similar results were obtained for the model combining the three regional features at five voxel spacings. Our study revealed the predictive role of the developed methods with multi-regional features for the preoperative assessment of prognostic factors in NSCLC. The analysis of different voxel spacing and model interpretability strengthens the experimental findings and contributes to understanding the biological significance of the radiological phenotype.

A Novel MBAS-RF Approach to Predict Mechanical Properties of Geopolymer-Based Compositions.

Using gels to replace a certain amount of cement in concrete is conducive to the green concrete industry, while testing the compressive strength (CS) of geopolymer concrete requires a substantial amount of substantial effort and expense. To solve the above issue, a hybrid machine learning model of a modified beetle antennae search (MBAS) algorithm and random forest (RF) algorithm was developed in this study to model the CS of geopolymer concrete, in which MBAS was employed to adjust the hyperparameters of the RF model. The performance of the MBAS was verified by the relationship between 10-fold cross-validation (10-fold CV) and root mean square error (RMSE) value, and the prediction performance of the MBAS and RF hybrid machine learning model was verified by evaluating the correlation coefficient (R) and RMSE values and comparing with other models. The results show that the MBAS can effectively tune the performance of the RF model; the hybrid machine learning model had high R values (training set R = 0.9162 and test set R = 0.9071) and low RMSE values (training set RMSE = 7.111 and test set RMSE = 7.4345) at the same time, which indicated that the prediction accuracy was high; NaOH molarity was confirmed as the most important parameter regarding the CS of geopolymer concrete, with the importance score of 3.7848, and grade 4/10 mm was confirmed as the least important parameter, with the importance score of 0.5667.

Prospective Detection of Early Lung Cancer in Patients With COPD in Regular Care by Electronic Nose Analysis of Exhaled Breath

Patients with COPD are at high risk of lung cancer developing, but no validated predictive biomarkers have been reported to identify these patients. Molecular profiling of exhaled breath by electronic nose (eNose) technology may qualify for early detection of lung cancer in patients with COPD. Can eNose technology be used for prospective detection of early lung cancer in patients with COPD? BreathCloud is a real-world multicenter prospective follow-up study using diagnostic and monitoring visits in day-to-day clinical care of patients with a standardized diagnosis of asthma, COPD, or lung cancer. Breath profiles were collected at inclusion in duplicate by a metal-oxide semiconductor eNose positioned at the rear end of a pneumotachograph (SpiroNose; Breathomix). All patients with COPD were managed according to standard clinical care, and the incidence of clinically diagnosed lung cancer was prospectively monitored for 2 years. Data analysis involved advanced signal processing, ambient air correction, and statistics based on principal component (PC) analysis, linear discriminant analysis, and receiver operating characteristic analysis. Exhaled breath data from 682 patients with COPD and 211 patients with lung cancer were available. Thirty-seven patients with COPD (5.4%) demonstrated clinically manifest lung cancer within 2 years after inclusion. Principal components 1, 2, and 3 were significantly different between patients with COPD and those with lung cancer in both training and validation sets with areas under the receiver operating characteristic curve of 0.89 (95% CI, 0.83-0.95) and 0.86 (95% CI, 0.81-0.89). The same three PCs showed significant differences (P< .01) at baseline between patients with COPD who did and did not subsequently demonstrate lung cancer within 2 years, with a cross-validation value of 87%and an area under the receiver operating characteristic curve of 0.90 (95% CI, 0.84-0.95). Exhaled breath analysis by eNose identified patients with COPD in whom lung cancer became clinically manifest within 2 years after inclusion. These results show that eNose assessment may detect early stages of lung cancer in patients with COPD.

Proteomics and cytokine analyses distinguish myalgic encephalomyelitis/chronic fatigue syndrome cases from controls

BackgroundMyalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) is a complex, heterogenous disease characterized by unexplained persistent fatigue and other features including cognitive impairment, myalgias, post-exertional malaise, and immune system dysfunction. Cytokines are present in plasma and encapsulated in extracellular vesicles (EVs), but there have been only a few reports of EV characteristics and cargo in ME/CFS. Several small studies have previously described plasma proteins or protein pathways that are associated with ME/CFS.MethodsWe prepared extracellular vesicles (EVs) from frozen plasma samples from a cohort of Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) cases and controls with prior published plasma cytokine and plasma proteomics data. The cytokine content of the plasma-derived extracellular vesicles was determined by a multiplex assay and differences between patients and controls were assessed. We then performed multi-omic statistical analyses that considered not only this new data, but extensive clinical data describing the health of the subjects.ResultsME/CFS cases exhibited greater size and concentration of EVs in plasma. Assays of cytokine content in EVs revealed IL2 was significantly higher in cases. We observed numerous correlations among EV cytokines, among plasma cytokines, and among plasma proteins from mass spectrometry proteomics. Significant correlations between clinical data and protein levels suggest roles of particular proteins and pathways in the disease. For example, higher levels of the pro-inflammatory cytokines Granulocyte-Monocyte Colony-Stimulating Factor (CSF2) and Tumor Necrosis Factor (TNFα) were correlated with greater physical and fatigue symptoms in ME/CFS cases. Higher serine protease SERPINA5, which is involved in hemostasis, was correlated with higher SF-36 general health scores in ME/CFS. Machine learning classifiers were able to identify a list of 20 proteins that could discriminate between cases and controls, with XGBoost providing the best classification with 86.1% accuracy and a cross-validated AUROC value of 0.947. Random Forest distinguished cases from controls with 79.1% accuracy and an AUROC value of 0.891 using only 7 proteins.ConclusionsThese findings add to the substantial number of objective differences in biomolecules that have been identified in individuals with ME/CFS. The observed correlations of proteins important in immune responses and hemostasis with clinical data further implicates a disturbance of these functions in ME/CFS.

PEMODELAN HARGA SAHAM PERUSAHAAN PROPERTI DAN REAL ESTATE MENGGUNAKAN REGRESI LONGITUDINAL SPLINE TRUNCATED DILENGKAPI GUI R

Stocks are one of the most popular financial instruments traded in the capital market. One of stock prices fluctuate up and down due to the influence of several factors, one of which is inflation. Stocks in the property and real estate sectors are important indicators to determine the level of a country economy. Data on several stock prices is one case of longitudinal data in economic field. The data is divided into 2 parts, namely in sample data from January 2016 to October 2020 and out sample data from November 2020 to December 2021. In this study, longitudinal stock price data is modeling using nonparametric spline truncated. The best spline truncated model is determined by the order and the optimal number of knot points based on the minimum Generalized Cross Validation value. Spline truncated nonparametric regression modeling for longitudinal data in this study is equipped with Graphical User Interface (GUI) that can facilitate the data processing. The results of the analysis show that the best longitudinal spline truncated regression model obtained on 2nd order with 5 knot points. 95.04% value of indicates the model is a strong model. In the evaluation of the best model, the MAPE data out sample value is 16.45%. It indicates the model has good forecasting ability.

Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM2.5 Using a Random Forest Approach

The latest epidemiological studies have revealed that the adverse health effects of PM2.5 have impacts beyond respiratory and cardio-vascular diseases and also affect the development of the brain and metabolic diseases. The need for accurate and spatio-temporally resolved PM2.5 data has thus been substantiated. While the selective information provided by station measurements is mostly insufficient for area-wide monitoring, satellite data have been increasingly applied to comprehensively monitor PM2.5 distributions. Although the accuracy and reliability of satellite-based PM2.5 estimations have increased, most studies still rely on a single sensor. However, several datasets have become available in the meantime, which raises the need for a systematic analysis. This study presents the first systematic evaluation of four satellite-based AOD datasets obtained from different sensors and retrieval methodologies to derive ground-level PM2.5 concentrations. We apply a random forest approach and analyze the effect of the resolution and coverage of the satellite data and the impact of proxy data on the performance. We examine AOD data from the Moderate resolution Imaging spectroradiometer (MODIS) onboard Terra and Aqua satellites, including Dark Target (DT) algorithm products and the Multi-Angle Implementation of Atmospheric Correction (MAIAC) product. Additionally, we explore more recent datasets from the Sea and Land Surface Temperature Radiometer (SLSTR) onboard Sentinel-3a and from the Tropospheric Monitoring Instrument (TROPOMI) operating on the Sentinel-5 precursor (S5p). The method is demonstrated for Germany and the year 2018, where a dense in situ measurement network and relevant proxy data are available. Overall, the model performance is satisfactory for all four datasets with cross-validated R2 values ranging from 0.68 to 0.77 and excellent for MODIS AOD reaching correlations of almost 0.9. We find a strong dependency of the model performance on the coverage and resolution of the AOD training data. Feature importance rankings show that AOD has less weight compared to proxy data for SLSTR and TROPOMI.

MP17-11 REFINEMENT AND EXTERNAL VALIDATION OF A NOVEL MULTIPLEX URINE TEST FOR HIGH-GRADE PROSTATE CANCER

MP17-11 REFINEMENT AND EXTERNAL VALIDATION OF A NOVEL MULTIPLEX URINE TEST FOR HIGH-GRADE PROSTATE CANCER

Profiling the structural determinants of pyrrolidine derivative as gelatinases (MMP-2 and MMP-9) inhibitors using in silico approaches

Quantitative structure activity relationship (QSAR) studies on pyrrolidine derivatives have been established using CoMFA, CoMSIA, and Hologram QSAR analysis to estimate the values (pIC50) of gelatinase inhibitors. When the CoMFA cross-validation value, Q², was 0.625, the training set coefficient of determination, R² was 0.981. In CoMSIA, Q² was 0.749 and R² was 0.988. In the HQSAR, Q² was 0.84 and R² was 0.946. Visualization of these models was performed by contour maps showing favorable and unfavorable regions for activity, while visualization of HQSAR model was performed by a colored atomic contribution graph. Based on the results obtained of external validation, the CoMSIA model was statistically more significant and robust and was selected as the best model to predict new, more active inhibitors. To study the modes of interactions of the predicted compounds in the active site of MMP-2 and MMP-9, a simulation of molecular docking was realized. A combined study of MD simulations and calculation of free binding energy, were also carried out to validate the results obtained on the best predicted and most active compound in dataset and the compound NNGH as control compound. The results confirm the molecular docking results and indicate that the predicted ligands were stable in the binding site of MMP-2 and MMP-9.

Nonparametric Regression Modeling with Fourier Series Approach on Poverty Cases in West Sumatra Province

Poverty is a complex problem that has an impact on various social problems such as education, unemployment, health and economic growth. Therefore, the problem of poverty is important to overcome in order to create population welfare. One of the analyses that can be used to model the percentage of poverty is regression analysis. Regression analysis is divided into two approaches, namely parametric and nonparametric. Parametric regression has several assumptions while, the only assumption nonparametric regression shape of the curve does not form a certain pattern. There are several approaches to nonparametric regression, one of which is the Fourier Series. The purpose of this study is to model the percentage of poverty in West Sumatra Province. The unclear shape of the curve in the data used is a consideration for using nonparametric regression. Then it is known that the data used in this study is data per region which tends to have a fluctuating nature. So it is suitable to use the Fourier series approach. In this research, nonparametric regression modeling with one, two, and three oscillation parameters was attempted. The best model was obtained which consisted of two oscillation parameters with a Generalized Cross Validation (GCV) value of 2.110 and R² of 92.44%.

Price Prediction of Ames Housing Through Advanced Regression Techniques

House price fluctuates at all times, and it is necessary to predict the future house prices for the reason that it is crucial for buyers, property investors and the economy as a whole. This article reveals the important features relating to house price and demonstrates the method of predicting house prices in Ames, Iowa through advanced regression techniques, such as correlation, feature engineering, and model building. The prediction is performed through machine learning methods based on the dataset containing almost all the features of residential houses in Ames, Iowa. The dataset was originally splitted into the ‘train set’ and the ‘test set’. Six models were chosen, LASSO Regression, Elastic Net Regression, Gradient Boosting Regression, XBoost, LightGBM, and Stacked Model, for evaluation under the five-fold cross-validation to check the performance of the algorithms and report the root-mean-square logarithmic error. Eventually, Gradient Boosting Regression Model is chosen by the lowest root-mean-square logarithmic error of 0.04615 and a low cross-validated value of 0.1174 to predict the house price to benefit buyers, investors, and the whole society.

Tropical Forest Top Height by GEDI: From Sparse Coverage to Continuous Data

Estimating consistent large-scale tropical forest height using remote sensing is essential for understanding forest-related carbon cycles. The Global Ecosystem Dynamics Investigation (GEDI) light detection and ranging (LiDAR) instrument employed on the International Space Station has collected unique vegetation structure data since April 2019. Our study shows the potential value of using remote-sensing (RS) data (i.e., optical Sentinel-2, radar Sentinel-1, and radar PALSAR-2) to extrapolate GEDI footprint-level forest canopy height model (CHM) measurements. We show that selected RS features can estimate vegetation heights with high precision by analyzing RS data, spaceborne GEDI LiDAR, and airborne LiDAR at four tropical forest sites in South America and Africa. We found that the GEDI relative height (RH) metric is the best at 98% (RH98), filtered by full-power shots with a sensitivity greater than 98%. We found that the optical Sentinel-2 indices are dominant with respect to radar from 77 possible features. We proposed the nine essential optical Sentinel-2 and the radar cross-polarization HV PALSAR-2 features in CHM estimation. Using only ten optimal indices for the regression problems can avoid unimportant features and reduce the computational effort. The predicted CHM was compared to the available airborne LiDAR data, resulting in an error of around 5 m. Finally, we tested cross-validation error values between South America and Africa, including around 40% from validation data in training to obtain a similar performance. We recommend that GEDI data be extracted from all continents to maintain consistent performance on a global scale. Combining GEDI and RS data is a promising method to advance our capability in mapping CHM values.

Using Transfer Learning of Convolutional Neural Network on Neck Radiographs to Identify Acute Epiglottitis.

Acute epiglottitis (AE) is a life-threatening condition and needs to be recognized timely. Diagnosis of AE with a lateral neck radiograph yields poor reliability and sensitivity. Convolutional neural networks (CNN) are powerful tools to assist the analysis of medical images. This study aimed to develop an artificial intelligence model using CNN-based transfer learning to identify AE in lateral neck radiographs. All cases in this study are from two hospitals, a medical center, and a local teaching hospital in Taiwan. In this retrospective study, we collected 251 lateral neck radiographs of patients with AE and 936 individuals without AE. Neck radiographs obtained from patients without and with AE were used as the input for model transfer learning in a pre-trained CNN including Inception V3, Densenet201, Resnet101, VGG19, and Inception V2 to select the optimal model. We used five-fold cross-validation to estimate the performance of the selected model. The confusion matrix of the final model was analyzed. We found that Inception V3 yielded the best results as the optimal model among all pre-train models. Based on the average value of the fivefold cross-validation, the confusion metrics were obtained: accuracy = 0.92, precision = 0.94, recall = 0.90, and area under the curve (AUC) = 0.96. Using the Inception V3-based model can provide an excellent performance to identify AE based on radiographic images. We suggest using the CNN-based model which can offer a non-invasive, accurate, and fast diagnostic method for AE in the future.

Geographically Weighted Panel Regression Modelling of Human Development Index Data in East Kalimantan Province in 2017-2020

Geographically Weighted Panel Regression (GWPR) model is a panel regression model applied on spatial data. This research applied Fixed Effect Model (FEM) on panel regression as the global model and GWPR as the local model for Human Development Index (HDI) regencies/municipalities in East Kalimantan Province data over the years 2017-2020. The aim of this research is to obtain the GWPR model of HDI data, as well as to acquire factors that influence it. The parameter of GWPR model was estimated on each observation location using the Weighted Least Square (WLS) method, namely Ordinary Least Square (OLS) with addition of spatial weighting. The spatial weighting on GWPR model was calculated using fixed bisquare, fixed tricube, adaptive bisquare and adaptive tricube. After the selection process, the optimum weighting function is adaptive tricube which provides a minimum Cross Validation (CV) value of 5.1419. Based on GWPR parameter testing, factors that affect HDI are local and diverse in each 10 regencies/municipalities in East Kalimantan Province. These factors are the labor force participation rate, number of health facilities, Gini ratio, population growth rate, open unemployment rate, poverty gap index and percentage of food expenditure. The coefficient of determination of the GWPR model obtains a value of 94.36% with the RMSE value of 0.1221.

Modeling of COVID-19 Cases in Indonesia with the Method of Geographically Weighted Regression

The COVID-19 pandemic has spread to all corners of the world, including Indonesia. Various factors affect the spread of COVID-19 cases in an area so that the government and the community can make prevention and control efforts so that this pandemic does not spread. This study aims to model the number of COVID-19 cases in Indonesia using the Geographically Weighted Regression (GWR) method, which develops a linear regression model. The GWR model uses weights based on the location of each observation so that the model is obtained for that location. Determine the weighting on the bandwidth. Optimum bandwidth selection is obtained by minimizing the value of Cross-Validation (CV). The GWR model using a fixed bisquare kernel weighting function has an optimum bandwidth of 0.999948 with a minimum CV value of 397.076.128 with a coefficient of determination R2 of 85.1 %. The results show that the number of positive cases positively correlates with the number of patients who died from COVID-19. In contrast, the number of recovered patients negatively correlates with the number of patients who died from COVID-19.