Prediction of maximum oxygen uptake with different machine learning methods by using submaximal data

Evapotranspiration (ET) is a fundamental element in the hydrological cycle and plays a vital role in simulating hydrological effects of climate change. Accurate estimation of reference evapotranspiration (ETo) plays a significant role in the field of hydrometeorology and agrometeorology. There are many direct or indirect methods that are employed for estimation of ETo. Furthermore, soft computing approaches are also used to solve complex problems of estimation in various disciplines. The application of the soft computing techniques for estimating evapotranspiration has only been limited to few approaches including artificial neural network, fuzzy logic, and genetic algorithm. This study is conducted to apply non-conventional soft computing approaches for estimating ETo which include single decision tree (SDT), tree boost (TB), and decision tree forest (DTF). Monthly meteorological data spanning over 30 years (1987 to 2016) for six different cities located in arid, semi-arid, and humid regions of Pakistan is used for the estimation of ETo in this study. Seventeen input combinations comprising of various climatic parameters were developed to evaluate the impact of different parameters. A total of 306 models were developed using the SDT, TB, and DTF approaches for all six selected cities of Pakistan. The results of the developed models are then compared with ETo calculated by the standard method of modified Penmen equation. The TB modeling technique is found to perform best for various cities located in different climatic zones of Pakistan. Only two climatic parameters, the mean temperature and wind velocity, are found to be sufficient for efficient ETo prediction. To further validate the results, the outcome of the study was applied on some other climatic regions located in the USA, New Zealand, and China. Finally, it is concluded that these non-conventional approaches can also be considered superior alternatives to the available methods of estimating ETo.

Identifying pollution sources and predicting urban air quality using ensemble learning methods

The purpose of this study is to develop new multilayer feed-forward artificial neural network (ANN)-based maximal oxygen uptake (VO2max) prediction models by using submaximal treadmill exercise and nonexercise data. Using 10-fold cross validation on the dataset, standard error of estimate (SEE) and multiple correlation coefficient (R) of the models are calculated. It is shown that the models including submaximal, standard nonexercise and questionnaire variables yield higher R and lower SEE than the ones including submaximal and standard nonexercise variables only. The results of ANN-based models are also compared with the ones obtained by Multiple Linear Regression (MLR) and Support Vector Machines (SVM). It is shown that ANN-based models perform better than MLR and SVM-based models for predicting VO2max.

The science of hydrological modeling has continuously evolved under the influence of rapid advancements in software and hardware technologies. Starting from simple rational formulae for estimating peak discharge and developing into sophisticated univariate predictive models, accurate conversion of rainfall into runoff and the assessment of inherent uncertainty has been a prime focus for researchers. Therefore, alternative data-driven methods have gained widespread attention in hydrology. Moreover, scientists often couple conventional machine learning models with data pre-processing techniques, i.e., wavelet transformation (WT), to enhance modelling accuracy. In this context, this research work attempts to explore the latent linkage between rainfall and runoff in Pothohar region of Pakistan by developing a novel linkage of five streamline techniques of machine learning, including single decision tree (SDT), decision tree forest (DTF), tree boost (TB), multilayer perceptron (MLP), and gene expression modeling (GEP), with a more sophisticated variant of WT, i.e., maximal overlap discrete wavelet transformation (MODWT), for boundary correction of the transformed components of timeseries data. This study also implements these machine learning models in a stand-alone mode for a more comprehensive comparative analysis of performances. Furthermore, the study uses a combined-basin approach that divides Pothohar region into two basins to compensate for the complex topographic division of the study area. The results indicate that MODWT-based DTF outperformed other stand-alone and hybrid models in terms of modeling accuracy. In the first scenario, considering the Bunha-Kahan River basin, MODWT-DTF yielded the highest NSE (0.86) and the lowest RMSE (220.45 mm) and R2 (0.92 at lag order 3 (Lo3)) when transformed with daubechies4 (db4) at level three. While in the Soan-Haro River basin, MODWT-DTF produced the highest accuracy modeling at lag order 4 (Lo4) (NSE = 0.88, RMSE = 21.72 m3/s, and R2 = 0.91). The highly accurate performance of 3- and 4-days lagged models reflects the temporal consistency in hydrological response of the study area. The comparison of simple and hybrid model performance indicates up to a 55% increase in modeling accuracy due to data pre-processing with wavelet transformation.

In our earlier study (Hiilloskorpi et al. Int J Sports Med 1999;:438–443) we developed a prediction equation to estimate energy expenditure kcal*min−1 (EE) during exercise, based on the linear relationship with heart rate (HR) and EE during intensive activity. The aim of our new study was to develop a prediction equation which is applicable also for all daily activities. In our earlier study (Hiilloskorpi et al. Int J Sports Med 1999; 20:438–443) we developed a prediction equation to estimate energy expenditure (kcal*min−1 (EE) during exercise, based on the linear relationship with heart rate (HR) and EE during intensive activity. The aim of our new study was to develop a prediction equation which is applicable also for all daily activities. In this presentation we assessed possibilities to improve the EE estimation using three different HR variables as a predictors: HR alone, % of HR reserve [100*(activity HR - resting HR) / (maximal HR - resting HR)], and the difference between activity HR and resting HR. The study group consisted of 89 healthy volunteers, 42 men (mean age 36.5[sd 7.6]y., BMI 24.5 [2.4] kg*m−2, VO2 max 45.2 [6.5]ml*kg*min−1) and 47 women (37.5 [9.5]y., 23.3 [3.4] kg*m−2, 36.3 [5.4]ml*kg*min−1). The subjects performed an exercise test consisting two different parts: a low-activity part, during which the subjects moved lightweight objects while sitting and standing, during the second part the subjects did a maximal treadmill uphill-walking test (Lasset, Finland, 1986). Oxygen uptake VO2 (l*min−1) and carbon dioxide production (VCO2, l*min−1) were calculated from the gas analyses done by indirect calorimetry (SensorMedics Vmax 29 (California, USA). HR was registered by electrocardiography (Max-1, Marquette Electronics (Wisconsin, USA). Generalised linear models were used for the prediction of EE. Repeated EE values of the tests were predicted by the corresponding HR values from the same test. The other predictors were body weight, sex and the intensity of exercise. When compared to the measured EE values the standard error of estimate (SEE) was 1.41 kcal*min−1, when HR was used as a predictor in the equation. The best EE prediction was when % of HR reserve was used instead of HR, the SEE was then 1.02 kcal*min−1. The SEE was 1.09 kcal*min−1 when the difference between activity HR and resting HR was used in the prediction equation. Supported by Polar Electro Oy and The Ministry of Health, Finland.

to compare the predictive accuracy of regression models for a non-consecutive day VO2 max/Wingate testing protocol and a same day testing protocol. Participants (N=23) completed a treadmill GXT and Wingate cycle test. Participants (n=12) completed testing on non-consecutive days (NON) and (n=11) the same day (SAME). VO2 max (L/min) and peak power (PP) were collected. Linear regression analysis of NON revealed R2=0.808 and prediction equation Ŷ=1.499+0.004X and SAME showed R2=0.861 and prediction equation Ŷ=1.407+0.003X. NON standard error of estimate (SEE) and standard error of estimate percent (SEE %) were 0.62 L/min and 15.23%, respectively. SAME SEE and SEE% were 0.34 L/min and 10.98%, respectively. These results indicate PP obtained the same day of VO2max testing is a better predictor of cardiorespiratory fitness. Similarly, the SAME model is more accurate according to SEE and SEE%. This may be due to diminished effects of training adaptations that could occur 2-7 days between testing sessions during the NON testing protocol in healthy, active young adults.

Machine Learning (ML) provides a theoretical and methodological framework that allows to quantify the relationship between the user's Quality of Experience (OoE) and the network's Quality of Service (QoS). In the literature, several ML-based QoS/QoE correlation models have been proposed. All of those models use inductive supervised learning techniques and most of them are built in an offline batch manner using different ML methods such as: Least Squares Regression, Artificial Neural Netwcorks, Naive Bayes classifier, Support Vector Machines, k-Nearest Neighbors, Decision Trees, and Random Forest. This paper aims to evaluate these different ML methods and determine the most suitable one for the task of establishing the QoS/QoE correlation. The comparisons show that the Decision Trees and Random Forest models give the best results to this end.

BackgroundPancreatic cancer is the fourth leading cause of cancer death in the United States. Consequently, identification of clinically relevant biomarkers for the early detection of this cancer type is urgently needed. In recent years, proteomics profiling techniques combined with various data analysis methods have been successfully used to gain critical insights into processes and mechanisms underlying pathologic conditions, particularly as they relate to cancer. However, the high dimensionality of proteomics data combined with their relatively small sample sizes poses a significant challenge to current data mining methodology where many of the standard methods cannot be applied directly. Here, we propose a novel methodological framework using machine learning method, in which decision tree based classifier ensembles coupled with feature selection methods, is applied to proteomics data generated from premalignant pancreatic cancer.ResultsThis study explores the utility of three different feature selection schemas (Student t test, Wilcoxon rank sum test and genetic algorithm) to reduce the high dimensionality of a pancreatic cancer proteomic dataset. Using the top features selected from each method, we compared the prediction performances of a single decision tree algorithm C4.5 with six different decision-tree based classifier ensembles (Random forest, Stacked generalization, Bagging, Adaboost, Logitboost and Multiboost). We show that ensemble classifiers always outperform single decision tree classifier in having greater accuracies and smaller prediction errors when applied to a pancreatic cancer proteomics dataset.ConclusionIn our cross validation framework, classifier ensembles generally have better classification accuracies compared to that of a single decision tree when applied to a pancreatic cancer proteomic dataset, thus suggesting its utility in future proteomics data analysis. Additionally, the use of feature selection method allows us to select biomarkers with potentially important roles in cancer development, therefore highlighting the validity of this method.

Patient-targeted therapies have recently been highlighted as important. An important development in the treatment of metastatic non-small cell lung cancer (NSCLC) has been the tailoring of therapy on the basis of histology. A pathology diagnosis of “non-specified NSCLC” is no longer routinely acceptable; an effective approach for classification of adenocarcinoma (AC) and squamous carcinoma (SC) histotypes is needed for optimizing therapy. In this study, we present a robust and objective automatic classification system for real time classification of AC and SC based on morphological tissue pattern of H&E images alone to assist medical experts in diagnosis of lung cancer. Various original and extended Densitometric and Haralick’s texture features are used to extract image features, and a Boosting algorithm is utilized to train the classifier, together with alternative decision tree as the base learner. For evaluation, 369 tissue samples were collected in tissue microarray format, including 97 adenocarcinoma and 272 squamous carcinoma samples. Using 10-fold cross validation, the technique achieved high accuracy of 92.41%, and we also found that the two Boosting algorithms (cw-Boost and AdaBoost.M1) perform consistently well in comparison with other popularly adopted machine learning methods, including support vector machine, neural network, single decision tree and alternative decision tree. This approach offers a robust, objective and rapid procedure for optimized patient-targeted therapies.Keywordsmorphological classificationcomputer visionadenocarcinomasquamous carcinomaboostingtissue microarray

The purpose of this investigation was to cross-validate existing VO2max prediction equations on samples of aerobically trained males and females. A total of 142 aerobically trained males (mean +/- SD; 39.0 +/- 11.1 yr, N = 93) and females (39.7 +/- 10.1 yr, N = 49) performed a maximal incremental test to determine actual VO2max on a cycle ergometer. The predicted VO2max values from 18 equations (nine for each gender) were compared with actual VO2max by examining the constant error (CE), standard error of estimate (SEE), correlation coefficient (r), and total error (TE). The results of this investigation indicated that all of the equations resulted in significant (P < 0.006) CE values ranging from -216 to 1415 mL x min(-1) for the males and 132 to 1037 mL x min(-1) for the females. In addition the SEE, r, and TE values ranged from 266 to 609 mL x min(-1), 0.36 to 0.88, and 317 to 1535 mL x min(-1), respectively. Furthermore, the lowest TE values for the males and females represented 10% and 12% of the mean actual VO2max values, respectively. The results of the analysis indicated that the two equations using age, body weight, and the power output achieved at VO2 as predictor variables had the lowest SEE (7.7-9.8% of actual VO2max) and TE (10-12% of actual VO2max) values and are recommended for estimating VO2max in aerobically trained males and females. The magnitude of the TE values (>or= 20% of actual VO2max) associated with the remaining 16 equations, however, were too large to be of practical value for estimating VO2max.

Industrial robots are widely used in critical tasks such as assembly, welding, and material handling as core components of modern manufacturing systems. For the reliable operation of these systems, early and accurate detection of execution failures is crucial. In this study, a comprehensive comparison of machine learning and deep learning methods is conducted for the classification of robot execution failures using data acquired from force-torque sensors. Three different feature engineering approaches are proposed. The first is a Baseline approach that includes 90 raw time-series features. The second is the Domain-6 approach, which consists of 6 basic statistical features per sensor (36 in total). The third is the Domain-12 approach, which comprises 12 comprehensive statistical features per sensor (72 in total). The domain features include the mean, standard deviation, minimum, maximum, range, slope, median, skewness, kurtosis, RMS, energy, and IQR. In total, ten classification algorithms are evaluated, including eight machine learning methods and two deep learning models: Support Vector Machines (SVM), Random Forest (RF), k-Nearest Neighbors (KNN), Artificial Neural Network (ANN), Naive Bayes (NB), Decision Trees (DT), eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM-LGBM), as well as a One-Dimensional Convolutional Neural Network (CNN-1D) and Long Short-Term Memory (LSTM). For traditional machine learning algorithms, 5 × 5 nested cross-validation is used, whereas for deep learning models, 5-fold cross-validation with a 20% validation split is employed. To ensure statistical reliability, all experiments are repeated over 30 independent runs. The experimental results demonstrate that feature engineering has a decisive impact on classification performance. In addition, regardless of the feature set, the highest accuracy (93.85% ± 0.90) is achieved by the Naive Bayes classifier using the Baseline features. The Domain-12 feature set provides consistent improvements across many algorithms, with substantial performance gains. The results are reported using accuracy, precision, recall, and F1-score metrics and are supported by confusion matrices. Finally, permutation feature importance analysis indicates that the skewness features of the Fx and Fy sensors are the most critical variables for failure detection. Overall, these findings show that time-domain statistical features offer an effective approach for robot failure classification.

Artificial intelligence has been applied in wildfire science and management since the 1990s, with early applications including neural networks and expert systems. Since then, the field has rapidly progressed congruently with the wide adoption of machine learning (ML) methods in the environmental sciences. Here, we present a scoping review of ML applications in wildfire science and management. Our overall objective is to improve awareness of ML methods among wildfire researchers and managers, as well as illustrate the diverse and challenging range of problems in wildfire science available to ML data scientists. To that end, we first present an overview of popular ML approaches used in wildfire science to date and then review the use of ML in wildfire science as broadly categorized into six problem domains, including (i) fuels characterization, fire detection, and mapping; (ii) fire weather and climate change; (iii) fire occurrence, susceptibility, and risk; (iv) fire behavior prediction; (v) fire effects; and (vi) fire management. Furthermore, we discuss the advantages and limitations of various ML approaches relating to data size, computational requirements, generalizability, and interpretability, as well as identify opportunities for future advances in the science and management of wildfires within a data science context. In total, to the end of 2019, we identified 300 relevant publications in which the most frequently used ML methods across problem domains included random forests, MaxEnt, artificial neural networks, decision trees, support vector machines, and genetic algorithms. As such, there exists opportunities to apply more current ML methods — including deep learning and agent-based learning — in the wildfire sciences, especially in instances involving very large multivariate datasets. We must recognize, however, that despite the ability of ML models to learn on their own, expertise in wildfire science is necessary to ensure realistic modelling of fire processes across multiple scales, while the complexity of some ML methods such as deep learning requires a dedicated and sophisticated knowledge of their application. Finally, we stress that the wildfire research and management communities play an active role in providing relevant, high-quality, and freely available wildfire data for use by practitioners of ML methods.

Developing Theory Using Machine Learning Methods

The aim of this research is to predict preplaced-aggregate concrete (PAC) compressive strength (CS) by using machine learning approaches such as gene expression programming (GEP) and random forest (RF). PAC requires injecting a portland cement-sand grout with admixtures into a mold after coarse aggregate has been deposited, making CS prediction complicated and requiring substantial study. Machine learning methods were used to cut down on the time and money needed for extensive experimental testing. The database includes 135 values for CS with eleven input variables. There is an acceptable degree of agreement between predicted and experimental values, as shown by the CS R2 values of 0.94 for GEP and 0.96 for RF. When comparing RF with GEP, RF performed better as measured by R2. The lower values displayed by the statistical error also showed that RF performed better than GEP. To compare, the GEP model's COV, MAE, RSME, and RMSLE were 0.527, 1.569, 2.706, and 0.133, whereas those for RF were 0.450, 1.648, 2.17, and 0.092. The SHAP analysis showed the effects of each input parameter, illuminating the positive effect of increasing the superplasticizer content on strength and the negative effect of raising the water-to-binder ratio. Using machine learning approaches to forecast the CS of PAC, this study has the potential to boost environmental protection and economic advantage.

Background: Dihydropteridone derivatives represent a novel class of PLK1 inhibitors, exhibiting promising anticancer activity and potential as chemotherapeutic drugs for glioblastoma. Objective: The aim of this study is to develop 2D and 3D-QSAR models to validate the anticancer activity of dihydropteridone derivatives and identify optimal structural characteristics for the design of new therapeutic agents. Methods: The Heuristic method (HM) was employed to construct a 2D-linear QSAR model, while the gene expression programming (GEP) algorithm was utilized to develop a 2D-nonlinear QSAR model. Additionally, the CoMSIA approach was introduced to investigate the impact of drug structure on activity. A total of 200 novel anti-glioma dihydropteridone compounds were designed, and their activity levels were predicted using chemical descriptors and molecular field maps. The compounds with the highest activity were subjected to molecular docking to confirm their binding affinity. Results: Within the analytical purview, the coefficient of determination (R2) for the HM linear model is elucidated at 0.6682, accompanied by an R2 cv of 0.5669 and a residual sum of squares (S2) of 0.0199. The GEP nonlinear model delineates coefficients of determination for the training and validation sets at 0.79 and 0.76, respectively. Empirical modeling outcomes underscore the preeminence of the 3D-QSAR model, succeeded by the GEP nonlinear model, whilst the HM linear model manifested suboptimal efficacy. The 3D paradigm evinced an exemplary fit, characterized by formidable Q2 (0.628) and R2 (0.928) values, complemented by an impressive F-value (12.194) and a minimized standard error of estimate (SEE) at 0.160. The most significant molecular descriptor in the 2D model, which included six descriptors, was identified as "Min exchange energy for a C-N bond" (MECN). By combining the MECN descriptor with the hydrophobic field, suggestions for the creation of novel medications were generated. This led to the identification of compound 21E.153, a novel dihydropteridone derivative, which exhibited outstanding antitumor properties and docking capabilities. Conclusion: The development of 2D and 3D-QSAR models, along with the innovative integration of contour maps and molecular descriptors, offer novel concepts and techniques for the design of glioblastoma chemotherapeutic agents.