XGBoost Machine-learning Model Research Articles

Identification and Validation of Anoikis-Related Signatures for Predicting Prognosis in Lung Adenocarcinoma with Machine Learning.

Lung adenocarcinoma (LUAD) is an aggressive cancer that has an extremely poor prognosis. As well as facilitating the detachment of cancer cells from the primary tumor site, anoikis plays an important role in cancer metastasis. Few studies to date, however, have examined the role of anoikis in LUAD, in patient prognosis. A total of 316 anoikis-related genes (ANRGs) integrated from Genecards and Harmonizome portals. LUAD transcriptome data were retrieved from the Genotype-Tissue Expression Project (GEO) and The Cancer Genome Atlas (TCGA). Anoikis-related prognostic genes (ANRGs) were primarily screened by univariate Cox regression. All ANRGs were included in the Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression model to construct the powerful prognostic signature. This signature was validated and assessed using the Kaplan-Meier method as well as univariate and multivariate Cox regression analyses. Anoikis-related regulators of risk score were identified using a XG-boost machine learning model. The expression of ITGB4 protein was examined in a ZhengZhou University (ZZU) tissue cohort by immunohistochemistry, and the potential mechanisms of action of ITGB4 in LUAD were explored by GO, KEGG, and ingenuity pathway analyses and by GSEA. A risk score signature was constructed based on eight ANRGs, with high risk scores found to closely correlate with unfavorable clinical features. ITGB4 expression may be associated with 5-year over survival, with immunohistochemistry showed that the expression of ITGB4 was higher in LUAD than in nontumor tissues. Enrichment analysis suggested that ITGB4 may promote LUAD development by targeting E2F, MYC, and oxidative phosphorylation signaling pathways. Our anoikis-related signature from RNA-seq data may be a novel prognostic biomarker in patients with LUAD. It may help physicians develop personalized LUAD treatments in clinical practice. Moreover, ITGB4 may affect the development of LUAD through the oxidative phosphorylation pathway.

Risk Factors of Hyperglycemia After Treatment With the AKT Inhibitor Ipatasertib in the Prostate Cancer Setting: A Machine Learning-Based Investigation.

Hyperglycemia is a major adverse event of phosphatidylinositol 3-kinase/AKT inhibitor class of cancer therapeutics. Machine learning (ML) methodologies can identify and highlight how explanatory variables affect hyperglycemia risk. Using data from clinical trials of the AKT inhibitor ipatasertib (IPAT) in the metastatic castrate-resistant prostate cancer setting, we trained an XGBoost ML model to predict the incidence of grade ≥2 hyperglycemia (HGLY ≥ 2). Of the 1,364 patients included in our analysis, 19.4% (n = 265) of patients had HGLY ≥2 events with a median time of first onset of 28 days (range, 0-753 days), and 30.0% (n = 221) of patients on an IPAT regimen had at least one HGLY ≥2 event compared with 7.0% (n = 44) of patients on placebo. An 11-variable XGBoost model predicted HGLY ≥2 events well with an AUROC of 0.83 ± 0.02 (mean ± standard deviation). Using SHapley Additive exPlanations analysis, we found IPAT exposure and baseline HbA1c levels to be the strongest predictors of HGLY ≥2, with additional predictivity of baseline measurements of fasting glucose, magnesium, and high-density lipoproteins. The findings support using patients' prediabetic status as a key factor for hyperglycemia monitoring and/or trial exclusion criteria. Additionally, the model and relationships between explanatory variables and HGLY ≥2 described herein can help identify patients at high risk for hyperglycemia and develop rational risk mitigation strategies.

Predicting Thermoelectric Power Plants Diesel/Heavy Fuel Oil Engine Fuel Consumption Using Univariate Forecasting and XGBoost Machine Learning Models

Monitoring and controlling thermoelectric power plants (TPPs) operational parameters have become essential to ensure system reliability, especially in emergencies. Due to system complexity, operating parameters control is often performed based on technical know-how and simplified analytical models that can result in limited observations. An alternative to this task is using time series forecasting methods that seek to generalize system characteristics based on past information. However, the analysis of these techniques on large diesel/HFO engines used in Brazilian power plants under the dispatch regime has not yet been well-explored. Therefore, given the complex characteristics of engine fuel consumption during power generation, this work aimed to investigate patterns generalization abilities when linear and nonlinear univariate forecasting models are used on a representative database related to an engine-driven generator used in a TPP located in Pernambuco, Brazil. Fuel consumption predictions based on artificial neural networks were directly compared to XGBoost regressor adaptation to perform this task as an alternative with lower computational cost. AR and ARIMA linear models were applied as a benchmark, and the PSO optimizer was used as an alternative during model adjustment. In summary, it was possible to observe that AR and ARIMA-PSO had similar performances in operations and lower error distributions during full-load power output with normal error frequency distribution of −0.03 ± 3.55 and 0.03 ± 3.78 kg/h, respectively. Despite their similarities, ARIMA-PSO achieved better adherence in capturing load adjustment periods. On the other hand, the nonlinear approaches NAR and XGBoost showed significantly better performance, achieving mean absolute error reductions of 42.37% and 30.30%, respectively, when compared with the best linear model. XGBoost modeling was 8.7 times computationally faster than NAR during training. The nonlinear models were better at capturing disturbances related to fuel consumption ramp, shut-down, and sudden fluctuations steps, despite being inferior in forecasting at full-load, especially XGBoost due to its high sensitivity with slight fuel consumption variations.

Prediction model for gestational diabetes mellitus using the XG Boost machine learning algorithm.

To develop the extreme gradient boosting (XG Boost) machine learning (ML) model for predicting gestational diabetes mellitus (GDM) compared with a model using the traditional logistic regression (LR) method. A case-control study was carried out among pregnant women, who were assigned to either the training set (these women were recruited from August 2019 to November 2019) or the testing set (these women were recruited in August 2020). We applied the XG Boost ML model approach to identify the best set of predictors out of a set of 33 variables. The performance of the prediction model was determined by using the area under the receiver operating characteristic (ROC) curve (AUC) to assess discrimination, and the Hosmer-Lemeshow (HL) test and calibration plots to assess calibration. Decision curve analysis (DCA) was introduced to evaluate the clinical use of each of the models. A total of 735 and 190 pregnant women were included in the training and testing sets, respectively. The XG Boost ML model, which included 20 predictors, resulted in an AUC of 0.946 and yielded a predictive accuracy of 0.875, whereas the model using a traditional LR included four predictors and presented an AUC of 0.752 and yielded a predictive accuracy of 0.786. The HL test and calibration plots show that the two models have good calibration. DCA indicated that treating only those women whom the XG Boost ML model predicts are at risk of GDM confers a net benefit compared with treating all women or treating none. The established model using XG Boost ML showed better predictive ability than the traditional LR model in terms of discrimination. The calibration performance of both models was good.

Prediction of wear of dental composite materials using machine learning algorithms

Since dental materials are worn down over time and eventually need to be replaced. Resin composites are frequently employed as dental restorative materials. By employing the in-vitro test findings of the pin-on-disc tribometer [ASTM G99-04], the goal of this study is to evaluate the capability of three different machine learning (ML) models in analyzing the wear of dental composite materials when immersed in chewable tobacco solution. Four distinct dental composite material samples are used in this investigation, and after being dipped in a chewing tobacco solution for a few days, the samples are taken out and subjected to a wear test. Three different ML models (MLP, KNN, XGBoost) have been chosen for predicting the wear of dental composite specimens. XGBoost ML model yields an R 2 value of 0.9996 and it performs noticeably better than the other approaches.

Induction Motor Fault Diagnosis Using Support Vector Machine, Neural Networks, and Boosting Methods

Induction motors are robust and cost effective; thus, they are commonly used as power sources in various industrial applications. However, due to the characteristics of induction motors, industrial processes can stop when motor failures occur. Thus, research is required to realize the quick and accurate diagnosis of faults in induction motors. In this study, we constructed an induction motor simulator with normal, rotor failure, and bearing failure states. Using this simulator, 1240 vibration datasets comprising 1024 data samples were obtained for each state. Then, failure diagnosis was performed on the acquired data using support vector machine, multilayer neural network, convolutional neural network, gradient boosting machine, and XGBoost machine learning models. The diagnostic accuracies and calculation speeds of these models were verified via stratified K-fold cross validation. In addition, a graphical user interface was designed and implemented for the proposed fault diagnosis technique. The experimental results demonstrate that the proposed fault diagnosis technique is suitable for diagnosing faults in induction motors.

Using a machine learning approach and big data to augment WASDE forecasts: Empirical evidence from US corn yield

AbstractThis paper investigates the accuracy of corn yield forecasts using machine learning with satellite and weather data. In addition, the study examines the incremental value of these forecasts to augment the World Agricultural Supply and Demand Estimates (WASDE) forecast. To illustrate the potential of machine learning methods for agricultural forecasting, publicly available data are collected from 1984 to 2021 for national corn yield, state corn yield, satellite variables, and weather variables and used with the XGBoost algorithm. The results show that the XGBoost model performed about the same but did not outperform the WASDE corn yield forecasts over a 12‐year out‐of‐sample period. The incremental value analysis results suggest that the XGBoost and WASDE forecasts capture similar information, and no incremental information exits. Although the XGBoost model does not outperform the WASDE August forecast, it is near real‐time and can be produced using publicly available data. The results indicate that the XGBoost machine learning models can produce reasonably accurate crop yield forecasts.

The association between hemoglobin A1c and all-cause mortality in the ICU: A cross-section study based on MIMIC-IV 2.0.

Hyperglycemia has been reported to be associated with the outcomes of patients in the intensive care unit (ICU). However, the relationship between hemoglobin A1c (HbA1c) and long-term or short-term mortality in the ICU is still unknown. This study used the Medical Information Mart for Intensive Care (MIMIC)-IV database to investigate the relationship between HbA1c and long-term or short-term mortality among ICU patients without a diabetes diagnosis. A total of 3,154 critically ill patients without a diabetes diagnosis who had HbA1c measurements were extracted and analyzed from the MIMIC-IV. The primary outcome was 1-year mortality, while the secondary outcomes were 30-day mortality and 90-day mortality after ICU discharge. HbA1c levels were classified into four levels according to three HbA1c values (5.0%, 5.7%, and 6.5%). The Cox regression model was used to investigate the relationship between the highest HbA1c measurement and mortality. Finally, this correlation was validated using the XGBoost machine learning model and Cox regression after propensity score matching (PSM). The study eventually included 3,154 critically ill patients without diabetes who had HbA1c measurements in the database. HbA1c levels of below 5.0% or above 6.5% were significantly associated with 1-year mortality after adjusting for covariates in Cox regression (HR: 1.37; 95% CI: 1.02-1.84 or HR: 1.62; 95% CI: 1.20-2.18). In addition, HbA1c 6.5% was linked to 30-day mortality (HR: 1.81; 95% CI: 1.21-2.71) and 90-day mortality (HR: 1.62; 95% CI: 1.14-2.29). The restricted cubic spline demonstrated a U-shaped relationship between HbA1c levels and 1-year mortality. The AUCs of the training and testing datasets in the XGBoost model were 0.928 and 0.826, respectively, while the SHAP plot revealed that HbA1c was somewhat important for the 1-year mortality. Higher HbA1c levels in Cox regression were still significantly associated with 1-year mortality after PSM for other factors. The 1-year mortality, 30-day mortality, and 90-day mortality rates for critically ill patients after discharge from ICU are significantly associated with HbA1c. HbA1c < 5.0% and ≥6.5% would increase 30-day, 90-day, and 1-year mortality, while levels between 5.0% and 6.5% of HbA1c did not significantly affect these outcomes.

Influence of Calcineurin Inhibitors and Genetic Polymorphism of Transporters on Enterohepatic Circulation and Exposure of Mycophenolic Acid in Chinese Adult Renal Allograft Recipients.

There is significant enterohepatic circulation (EHC) during the disposition of mycophenolic acid (MPA). The aim of this study was to elucidate factors influencing the EHC of MPA in Chinese adult renal allograft recipients. After 2weeks of therapy with mycophenolate mofetil or enteric-coated mycophenolate sodium, blood samples were collected from 125 patients at 0 to 12hours post-administration and MPA concentrations were determined. The influence of calcineurin inhibitors (CNIs) and genetic polymorphisms on MPA exposure and EHC was studied. The Shapley additive explanations method was used to estimate the impact of various factors on the area under the plasma drug concentration-time curve (AUC0-12h ) for MPA. An extreme gradient boosting (XGboost) machine learning-based model was established to predict AUC0-12h . Results showed that the dose-normalized AUC6-12h (dn-AUC6-12h ) of MPA was significantly lower in patients co-administered with cyclosporine (CsA) than in patients co-administered with tacrolimus (TAC) (P<.05). For patients co-administered with TAC, patients with ABCC2 C-24TCC or SLCO1B1 T521CTT genotypes had significantly higher values of dn-AUC6-12h (P<.05). Patients with SLCO1B3 334T/699G alleles had significantly lower dn-AUC6-12h values than homozygotes (P<.05). By introducing body weight, age, and EHC-related factors, including co-administered CNIs and genetic polymorphism of drug transporters, as covariates in the XGboost machine learning model, the prediction performance of AUC0-12h for MPA in Chinese adult renal allograft recipients can be improved.

Predicting mortality risk in dialysis: Assessment of risk factors using traditional and advanced modeling techniques within the Monitoring Dialysis Outcomes initiative.

Several factors affect the survival of End Stage Kidney Disease (ESKD) patients on dialysis. Machine learning (ML) models may help tackle multivariable and complex, often non-linear predictors of adverse clinical events in ESKD patients. In this study, we used advanced ML method as well as a traditional statistical method to develop and compare the risk factors for mortality prediction model in hemodialysis (HD) patients. We included data HD patients who had data across a baseline period of at least 1 year and 1 day in the internationally representative Monitoring Dialysis Outcomes (MONDO) Initiative dataset. Twenty-three input parameters considered in the model were chosen in an a priori manner. The prediction model used 1 year baseline data to predict death in the following 3 years. The dataset was randomly split into 80% training data and 20% testing data for model development. Two different modeling techniques were used to build the mortality prediction model. A total of 95,142 patients were included in the analysis sample. The area under the receiver operating curve (AUROC) of the model on the test data with XGBoost ML model was 0.84 on the training data and 0.80 on the test data. AUROC of the logistic regression model was 0.73 on training data and 0.75 on test data. Four out of the top five predictors were common to both modeling strategies. In the internationally representative MONDO data for HD patients, we describe the development of a ML model and a traditional statistical model that was suitable for classification of a prevalent HD patient's 3-year risk of death. While both models had a reasonably high AUROC, the ML model was able to identify levels of hematocrit (HCT) as an important risk factor in mortality. If implemented in clinical practice, such proof-of-concept models could be used to provide pre-emptive care for HD patients.

Abstract 11452: EHR-Based Machine Learning Model Predicts Drug-Induced QT Prolongation With Superior Performance Compared to Clinical Risk Predictors

Introduction: Many medications are associated with QTc prolongation (LQT). Current LQT predictors, including Tisdale and RISQ-PATH, have limited utility due to lack of generalizability to key use cases and modest model performance. A highly accurate LQT risk predictor utilizing data often available at the time of drug starts could broadly support LQT risk management. Hypothesis: An electronic health record (EHR)-based machine learning (ML) model is superior in predicting drug-induced LQT in a clinical population vs. Tisdale and RISQ-PATH. Methods: We identified all patients who began taking QT-prolonging medications (‘QTdrug’, per CredibleMeds) at Geisinger and had a follow-up 12-lead ECG within 1 year of medication start date while still on the drug. In case of multiple QTdrugs with overlapping dates, the last start date was used. Using 5-fold cross-validation, we trained an XGBoost ML model to predict LQT (QTc &gt;500 ms) within 1 year of QTdrug start date using EHR data as input, including 131 features across age, sex, smoking history, vitals, lab tests, comorbidities, baseline ECG metrics (within 3 years before drug start date) and QTdrug usage (ever on QTdrug, number of current QTdrugs). Results: QTdrug records were identified for 362,086 patients, and 6% had LQT within 1 year. The ML model demonstrated superior performance (Figure A) in predicting LQT (area under the curve (AUROC): 0.86, average precision score (AUPRC): 0.41) compared to RISQ-PATH (AUROC: 0.70, AUPRC: 0.19) and Tisdale (AUROC: 0.76, AUPRC: 0.21, calculated in a hospitalized subset N=113544). At the same specificity as RISQ-PATH (98%), the ML model had higher sensitivity (30% vs 10%) and positive predictive value (55% vs 29%). Similar comparisons were observed for Tisdale (Figure B). Conclusions: An EHR-based ML model outperforms commonly used risk calculators for predicting drug-induced LQT. This model can be used to stratify patients by risk of developing drug-induced LQT in a clinical setting.

The predictive value of pressure recording analytical method for the duration of mechanical ventilation in children undergoing cardiac surgery with an XGBoost-based machine learning model.

Prolonged mechanical ventilation in children undergoing cardiac surgery is related to the decrease in cardiac output. The pressure recording analytical method (PRAM) is a minimally invasive system for continuous hemodynamic monitoring. To evaluate the postoperative prognosis, our study explored the predictive value of hemodynamic management for the duration of mechanical ventilation (DMV). This retrospective study included 60 infants who underwent cardiac surgery. Cardiac index (CI), the maximal slope of systolic upstroke (dp/dtmax), and cardiac cycle efficiency (CCE) derived from PRAM were documented in each patient 0, 4, 8, and 12 h (T0, T1, T2, T3, and T4, respectively) after their admission to the intensive care unit (ICU). A linear mixed model was used to deal with the hemodynamic data. Correlation analysis, receiver operating characteristic (ROC), and a XGBoost machine learning model were used to find the key factors for prediction. Linear mixed model revealed time and group effect in CI and dp/dtmax. Prolonged DMV also have negative correlations with age, weight, CI at and dp/dtmax at T2. dp/dtmax outweighing CI was the strongest predictor (AUC of ROC: 0.978 vs. 0.811, p < 0.01). The machine learning model suggested that dp/dtmax at T2 ≤ 1.049 or < 1.049 in combination with CI at T0 ≤ 2.0 or >2.0 can predict whether prolonged DMV (AUC of ROC = 0.856). Cardiac dysfunction is associated with a prolonged DMV with hemodynamic evidence. CI measured by PRAM immediately after ICU admission and dp/dtmax 8h later are two key factors in predicting prolonged DMV.

Machine learning models including insulin resistance indexes for predicting liver stiffness in United States population: Data from NHANES.

BackgroundPrevention and treatment of liver fibrosis at an early stage is of great prognostic importance, whereas changes in liver stiffness are often overlooked in patients before the onset of obvious clinical symptoms. Recognition of liver fibrosis at an early stage is therefore essential.ObjectiveAn XGBoost machine learning model was constructed to predict participants' liver stiffness measures (LSM) from general characteristic information, blood test metrics and insulin resistance-related indexes, and to compare the fit efficacy of different datasets for LSM.MethodsAll data were obtained from the National Health and Nutrition Examination Survey (NHANES) for the time interval January 2017 to March 2020. Participants' general characteristics, Liver Ultrasound Transient Elastography (LUTE) information, indicators of blood tests and insulin resistance-related indexes were collected, including homeostasis model assessment of insulin resistance (HOMA-IR) and metabolic score for insulin resistance (METS-IR). Three datasets were generated based on the above information, respectively named dataset A (without the insulin resistance-related indexes as predictor variables), dataset B (with METS-IR as a predictor variable) and dataset C (with HOMA-IR as a predictor variable). XGBoost regression was used in the three datasets to construct machine learning models to predict LSM in participants. A random split was used to divide all participants included in the study into training and validation cohorts in a 3:1 ratio, and models were developed in the training cohort and validated with the validation cohort.ResultsA total of 3,564 participants were included in this study, 2,376 in the training cohort and 1,188 in the validation cohort, and all information was not statistically significantly different between the two cohorts (p > 0.05). In the training cohort, datasets A and B both had better predictive efficacy than dataset C for participants' LSM, with dataset B having the best fitting efficacy [±1.96 standard error (SD), (-1.49,1.48) kPa], which was similarly validated in the validation cohort [±1.96 SD, (-1.56,1.56) kPa].ConclusionsXGBoost machine learning models built from general characteristic information and clinically accessible blood test indicators are practicable for predicting LSM in participants, and a dataset that included METS-IR as a predictor variable would improve the accuracy and stability of the models.

Using Bayesian Optimization and Wavelet Decomposition in GPU for Arterial Blood Pressure Estimation.

Continuous monitoring of arterial blood pressure (ABP) of patients in hospital is currently carried out in an invasive way, which could represent a risk for them. In this paper, a noninvasive methodology to optimize ABP estimators using electrocardiogram and photoplethysmography signals is proposed. For this, the XGBoost machine learning model, optimized with Bayesian techniques, is executed in a Graphics Processing Unit, which drastically reduces execution time. The methodology is evaluated using the MIMIC-III Waveform Database. Systolic and diastolic pressures are estimated with mean absolute error values of 15.85 and 11.59 mmHg, respectively, similar to those of the state of the art. The main advantage of the proposed methodology with respect to others of the current state of the art is that it allows the optimization of the estimator model to be performed automatically and more efficiently at the computational level for the data available. Clinical Relevance- This approach has the advantage of using noninvasive methods to continuously monitor patient's arterial blood pressure, reducing the risk for patients.

Continuous on‐board behaviour classification using accelerometry: A case study with a new GPS‐3G‐Bluetooth system in Pacific black ducks

Abstract Over the past two decades, accelerometer (ACC) data have been increasingly used to study animal behaviours and energetics. However, the large amount of raw ACC data can be a burden to device storage and power consumption and in many cases may also require device retrieval for data collection. On‐board data processing to reduce data volume and power consumption for data transmission may hold promise to alleviate these problems and allow for next‐generation, smart trackers. We developed a tracking system processing raw ACC data on‐board of trackers into behaviours using an XGBoost machine learning model. We used this system on six free‐ranging Pacific black ducks Anas superciliosa to study eight behaviours every 2 s. The on‐board XGBoost model for behaviour classification had 92.04% overall accuracy. One day of continuous behaviour records was compressed on‐board to as little as 17.28 kB for routine transmission via the 3G network. We received behaviour data from the six ducks continuously for periods ranging between 56 days and 14 months. On‐board processing of raw ACC data and data transmission proofed highly energy efficient and came at a minimal weight cost to the trackers, providing great potential to open up new areas in ecological and behavioural research.

Prospective Clinical Validation of Virtual Patient-Specific Quality Assurance of Volumetric Modulated Arc Therapy Radiation Therapy Plans

Performing measurement-based patient-specific quality assurance (PSQA) is recognized as a resource-intensive and time inefficient task in the radiation therapy treatment workflow. Paired with technological refinements in modern radiation therapy, research toward measurement-free PSQA has seen increased interest during the past 5 years. However, these efforts have not been clinically implemented or prospectively validated in the United States. We propose a virtual QA (VQA) system and workflow to assess the safety and workload reduction of measurement-free PSQA. An XGBoost machine learning model was designed to predict PSQA outcomes of volumetric modulated arc therapy plans, represented as percent differences between the measured ion chamber point dose in a phantom and the corresponding planned dose. The final model was deployed within a web application to predict PSQA outcomes of clinical plans within an existing clinical workflow. The application also displays relevant feature importance and plan-specific distribution analyses relative to database plans for documentation and to aid physicist interpretation and evaluation. VQA predictions were prospectively validated over 3 months of measurements at our clinic to assess safety and efficiency gains. Over 3 months, VQA predictions for 445 volumetric modulated arc therapy plans were prospectively validated at our institution. VQA predictions for these plans had a mean absolute error of 1.08% ± 0.77%, with a maximum absolute error of 2.98%. Using a 1% prediction threshold (ie, plans predicted to have an absolute error <1% would not require a measurement) would yield a 69.2% reduction in QA workload, saving 32.5 hours per month on average, with 81.5% sensitivity, 72.4% specificity, and an area under the curve of 0.81 at a 3% clinical threshold and 100% sensitivity, 70% specificity, and an area under the curve of 0.93 at a 4% clinical threshold. This is the first prospective clinical implementation and validation of VQA in the United States, which we observed to be efficient. Using a conservative threshold, VQA can substantially reduce the number of required measurements for PSQA, leading to more effective allocation of clinical resources.

Satellite Derived Bathymetry with Sentinel-2 Imagery: Comparing Traditional Techniques with Advanced Methods and Machine Learning Ensemble Models

Accurate charting of nearshore bathymetry is critical to the safe and dependable use of coastal waterways frequented by the trading, fishing, tourism, and ocean energy industries. The accessibility of satellite imagery and the availability of various satellite-derived bathymetry (SDB) techniques have provided a cost-effective alternative to traditional in-situ bathymetric surveys. Furthermore, improved algorithms and the advancement of machine learning models have provided opportunity for higher quality bathymetric derivations. However, to date the relative accuracy and performance between traditional physics-based techniques, improved physics-based methods, and machine learning ensemble models have not been adequately quantified. In this study, nearshore bathymetry is derived from Sentinel-2 satellite imagery near La Parguera, Puerto Rico utilizing a traditional band-ratio algorithm, a band-ratio switching method, a random forest machine learning model, and the XGBoost machine learning model. The machine learning models returned comparable results and were markedly more accurate relative to other techniques; however, both machine learning models required an extensive training dataset. All models were constrained by environmental influences and image spatial resolution, which were assessed to be the limiting factors for routine use of satellite-derived bathymetry as a reliable method for hydrographic surveying.

Power load forecasting and interpretable models based on GS_XGBoost and SHAP

At present, the prediction accuracy of the power load prediction model is not high and the black box model prediction results are not explanatory, so this paper proposes an improved XGBoost based grid search for power load forecasting model (GS-XGBoost), and uses SHAP for model analysis and interpretation to show the impact of the feature set on the model in a visual way according to the marginal contribution of power load feature samples The impact of the prediction results is shown in a visual way to improve the interpretability of the model, and the key factors affecting the electricity load are identified based on the ranking results of the feature importance to provide decision support for the power system equipment maintenance plan, guide the formulation of strategies related to electricity production, resource procurement and pricing, and avoid market risks. In this paper, the electricity consumption dataset of an enterprise in Jiangsu is used as the research object, and the dataset is pre-processed with unique thermal coding, outlier analysis, null filling, and standardization. Through experimental comparison, the GS-XGBoost load forecasting model has the best prediction accuracy compared to SVR, random forest, decision tree, and XGBoost machine learning models, and then SHAP is applied to interpret the model and improve the model interpretability.

A Novel Machine Learning Predictive Tool Assessing Outpatient or Inpatient Designation for Medicare Patients Undergoing Total Knee Arthroplasty.

BackgroundRemoval of total joint arthroplasty from the inpatient-only list has created significant confusion regarding which patients qualify for an inpatient designation. The purpose of this study is to develop and validate a novel predictive tool for assessing who will be an outpatient vs inpatient after total knee arthroplasty (TKA).MethodsA cohort of Medicare patients undergoing primary TKA between January 2018 and September 2019 were retrospectively reviewed. Baseline demographics and patient characteristics were obtained, and their distributions for outpatient (less than 2 midnights) and inpatient stay were assessed. Subsequently, a XGBoost machine learning model was trained using 80% of the TKA patients, and the remaining 20% of patients were involved in testing the model's performance in terms of accuracy and the average area under the receive operating characteristic curve.ResultsEight hundred ninety-nine Medicare patients underwent TKA at our institution between January 2018 and September 2019. Of which, 625 patients had outpatients stays, and 274 qualified for inpatient designation. Significant associations were demonstrated between inpatient visits and the following factors: higher body mass index, increased age, better functional scores, multidimensional fatigue inventory, Charlson Comorbidity Index, American Society of Anesthesiologists score, female gender, cardiac history, and the Revised Cardiac Risk Index. The XGBoost model for predicting an inpatient or outpatient stay was 63.3% accurate, with area under the receive operating characteristic curve of 68.8%.ConclusionsUsing readily available key baseline characteristics, functional scores, and comorbidities, this machine-learning model accurately predicts the probability of an “outpatient” vs “inpatient” stay after TKA in the Medicare population. body mass index, age, VR12 functional scores, and multidimensional fatigue inventory scores had the highest influence on this predictive model.

Early Nutrient Diagnosis of Kentucky Bluegrass Combining Machine Learning and Compositional Methods

Kentucky bluegrass (Poa pratensis L.) is the most common perennial turfgrass species grown on playgrounds, municipal and residential lawn areas, and golf tees, fairways and roughs. Fertilization is the most efficient way to improve and maintain turfgrass aesthetic quality. Tissue diagnosis can guide fertilization, but tissue concentration ranges are biased by not taking into consideration nutrient inter-relationships, carryover effects and other key features. The centered log-ratio transformation reflects nutrient interactions in plants and avoids statistical biases. Machine learning (ML) models relate the target variable to the key features ex ante, and can predict future events from prior knowledge. The objective of his study was to predict turfgrass quality from key features and rank nutrients in the order of their limitations. The experimental setup comprised four N, three P, and four K rates applied on permanent plots during three consecutive years. Soils were a loam and an USGA sand. Eleven elements (N, S, P, K, Ca, Mg, B, Cu, Zn, Mn, Fe) were quantified in clippings collected during spring, summer and autumn every year. Turfgrass quality was categorized as target variable by color rating. Concentrations were centered log-ratioed (clr) partitioned into four quadrants in the confusion matrix generated by the xgboost ML model. The area under curve (AUC) and model accuracy were high to predict turfgrass color from the nutrient analyses of clippings collected in the preceding season, facilitating the seasonal adjustment of the fertilization regime to sustain high turfgrass quality. We provide a computational example to run the ML model and rank nutrients in the order of their limitations.