Large Number Of Input Variables Research Articles

A Machine Learning Model for Predicting In-Hospital Mortality in Chinese Patients With ST-Segment Elevation Myocardial Infarction: Findings From the China Myocardial Infarction Registry.

Machine learning (ML) risk prediction models, although much more accurate than traditional statistical methods, are inconvenient to use in clinical practice due to their nontransparency and requirement of a large number of input variables. We aimed to develop a precise, explainable, and flexible ML model to predict the risk of in-hospital mortality in patients with ST-segment elevation myocardial infarction (STEMI). This study recruited 18,744 patients enrolled in the 2013 China Acute Myocardial Infarction (CAMI) registry and 12,018 patients from the China Patient-Centered Evaluative Assessment of Cardiac Events (PEACE)-Retrospective Acute Myocardial Infarction Study. The Extreme Gradient Boosting (XGBoost) model was derived from 9616 patients in the CAMI registry (2014, 89 variables) with 5-fold cross-validation and validated on both the 9125 patients in the CAMI registry (89 variables) and the independent China PEACE cohort (10 variables). The Shapley Additive Explanations (SHAP) approach was employed to interpret the complex relationships embedded in the proposed model. In the XGBoost model for predicting all-cause in-hospital mortality, the variables with the top 8 most important scores were age, left ventricular ejection fraction, Killip class, heart rate, creatinine, blood glucose, white blood cell count, and use of angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin II receptor blockers (ARBs). The area under the curve (AUC) on the CAMI validation set was 0.896 (95% CI 0.884-0.909), significantly higher than the previous models. The AUC for the Global Registry of Acute Coronary Events (GRACE) model was 0.809 (95% CI 0.790-0.828), and for the TIMI model, it was 0.782 (95% CI 0.763-0.800). Despite the China PEACE validation set only having 10 available variables, the AUC reached 0.840 (0.829-0.852), showing a substantial improvement to the GRACE (0.762, 95% CI 0.748-0.776) and TIMI (0.789, 95% CI 0.776-0.803) scores. Several novel and nonlinear relationships were discovered between patients' characteristics and in-hospital mortality, including a U-shape pattern of high-density lipoprotein cholesterol (HDL-C). The proposed ML risk prediction model was highly accurate in predicting in-hospital mortality. Its flexible and explainable characteristics make the model convenient to use in clinical practice and could help guide patient management. ClinicalTrials.gov NCT01874691; https://clinicaltrials.gov/study/NCT01874691.

Odor Impact Assessment via Dispersion Model: Comparison of Different Input Meteorological Datasets

Dispersion modeling is a useful tool for reproducing the spatial–temporal distribution of pollutants emitted by industrial sites, particularly in the environmental odor field. One widely used tool, accepted by regulatory agencies for environmental impact assessments, is the CALPUFF model, which requires a large number of input variables, including meteorological and orographical variables. The reliability of model results depends on the accuracy of these input variables. The present research aims to discuss a comparative study of odor dispersion modeling by initializing the CALMET meteorological processor with different input data: surface and upper air observational meteorological data, 3D prognostic data, and a blend of prognostic and measured data. Two distinct sources (a point and an area source) and two different simulation domains in Cuba and Italy are considered. The analysis of results is based on odor impact criteria enforced in some Italian regions by computing the 98th percentile of odor peak concentrations on an annual basis. For the area source, simulation results reveal that the ‘OBS’ and ‘HYBRID’ modes are largely comparable, whereas prognostic data tend to underestimate the odor concentrations, likely due to a reduced percentage of wind calms. For point sources, different input meteorological settings provide comparable results, with no significant differences.

Utilizing nanotechnology and advanced machine learning for early detection of gastric cancer surgery

Nanotechnology has emerged as a promising frontier in revolutionizing the early diagnosis and surgical management of gastric cancers. The primary factors influencing curative efficacy in GIC patients are drug inefficacy and high surgical and pharmacological therapy recurrence rates. Due to its unique optical features, good biocompatibility, surface effects, and small size effects, nanotechnology is a developing and advanced area of study for detecting and treating cancer. Considering the limitations of GIC MRI and endoscopy and the complexity of gastric surgery, the early diagnosis and prompt treatment of gastric illnesses by nanotechnology has been a promising development. Nanoparticles directly target tumor cells, allowing their detection and removal. It also can be engineered to carry specific payloads, such as drugs or contrast agents, and enhance the efficacy and precision of cancer treatment. In this research, the boosting technique of machine learning was utilized to capture nonlinear interactions between a large number of input variables and outputs by using XGBoost and RNN-CNN as a classification method. The research sample included 350 patients, comprising 200 males and 150 females. The patients' mean ± SD was 50.34 ± 13.04 with a mean age of 50.34 ± 13.04. High-risk behaviors (P = 0.070), age at diagnosis (P = 0.034), distant metastasis (P = 0.004), and tumor stage (P = 0.014) were shown to have a statistically significant link with GC patient survival. AUC was 93.54%, Accuracy 93.54%, F1-score 93.57%, Precision 93.65%, and Recall 93.87% when analyzing stomach pictures. Integrating nanotechnology with advanced machine learning techniques holds promise for improving the diagnosis and treatment of gastric cancer, providing new avenues for precision medicine and better patient outcomes.

Machine-learning-based performance prediction of the energy pile heat pump system

The coefficient of performance (COP) is one of the important indicators for evaluating ground source heat pump systems, and a certain amount of machine learning algorithms have been applied for COP prediction. However, most of the current models use a large number of input variables, and it may be difficult to obtain all the required parameters for diagnosis of the sensors. Most of the influencing factors contain fluid temperature, which is not easy to measure, and there is a risk of leakage when sensors are deployed in the system. In this work, two-year field tests were conducted based on an energy pile heat pump system, where in situ results were used as sample points, and the more easily measured ambient temperature and humidity, room temperature and humidity, and hourly power consumption were used as input parameters. Spearman correlation analysis is used to rank the influencing factors, and the prediction of COP was developed by a hybrid artificial neural network (ANN), optimized via the particle swarm optimization algorithm (PSO), and finally compared with empirical models and their updates. Results showed that among the input variables, the most and least important factors affecting the COP are the ambient temperature and ambient humidity, respectively. The prediction results of PSO-ANN were in the best agreement with the measured values, and the accuracy was higher than that of the empirical regression models. The proposed method noted in this study can be used as a supplement to the current COP prediction.

Development of Symbolic Expressions Ensemble for Breast Cancer Type Classification Using Genetic Programming Symbolic Classifier and Decision Tree Classifier

Breast cancer is a type of cancer with several sub-types. It occurs when cells in breast tissue grow out of control. The accurate sub-type classification of a patient diagnosed with breast cancer is mandatory for the application of proper treatment. Breast cancer classification based on gene expression is challenging even for artificial intelligence (AI) due to the large number of gene expressions. The idea in this paper is to utilize the genetic programming symbolic classifier (GPSC) on the publicly available dataset to obtain a set of symbolic expressions (SEs) that can classify the breast cancer sub-type using gene expressions with high classification accuracy. The initial problem with the used dataset is a large number of input variables (54,676 gene expressions), a small number of dataset samples (151 samples), and six classes of breast cancer sub-types that are highly imbalanced. The large number of input variables is solved with principal component analysis (PCA), while the small number of samples and the large imbalance between class samples are solved with the application of different oversampling methods generating different dataset variations. On each oversampled dataset, the GPSC with random hyperparameter values search (RHVS) method is trained using 5-fold cross validation (5CV) to obtain a set of SEs. The best set of SEs is chosen based on mean values of accuracy (ACC), the area under the receiving operating characteristic curve (AUC), precision, recall, and F1-score values. In this case, the highest classification accuracy is equal to 0.992 across all evaluation metric methods. The best set of SEs is additionally combined with a decision tree classifier, which slightly improves ACC to 0.994.

A new approach to building surrogate models of high-fidelity stochastic simulations: PARIN (PARameter as Input-variable)

High-fidelity simulations are computationally expensive to evaluate for optimization and sensitivity analysis applications. One popular method to avoid this problem is utilizing surrogate models, which approximate the simulations with cheaper-to-evaluate functions. The existing surrogate modeling techniques are designed for deterministic systems, with only a few approaches available for stochastic simulations. This study proposes a new method, called PARIN (PARameter as INput), to efficiently construct accurate surrogate models of high-fidelity stochastic simulations. PARIN is compared to existing approaches in terms of accuracy and efficiency. The results reveal that PARIN generally has a lower normalized root mean square error in predicting the mean and standard deviation of the simulation outputs and that the output distribution predicted by PARIN has the lowest Wasserstein distance from the actual output distribution. However, both metrics for PARIN estimates deteriorate for simulations with a significantly large number of input variables in low computational-budget cases.

A study of Stall-Induced Vibrations using Surrogate-Based Optimization

Stall-Induced Vibrations (SIV) are an important design consideration for wind turbine blade design, especially for large, modern, wind turbines with highly flexible blades. Their severity depends on both the inflow and the structural characteristics of the blades. Studying SIV has a high computational cost, because it requires high-fidelity aeroelastic simulations, and potentially a large number of input variables. In an effort to reduce the computational cost of the domain exploration, in this work we have adopted a Surrogate-Based Optimization (SBO) framework. This way, the combination of input variables that lead to SIV can be explored with the minimum number of aeroelastic simulations. The proposed SBO framework can use any type of surrogate model, and leverages Delaunay triangulation to iteratively select samples to refine the surrogate model. The occurrence and severity of SIV on the IEA 10MW turbine is studied in a five variables space consisting of: wind speed, yaw angle, vertical wind shear, wind veer, and atmospheric temperature. A well-trained surrogate model is developed and used to predict the damping ratio of the first blade edgewise mode in the entire inflow space at a reduced computational cost. Sensitivity analysis of the predicted damping ratio shows that yaw angle is the most influential variable, while temperature is the least influential variable in terms of inflow conditions that can lead to the occurrence of SIV. Inflow conditions with a moderate yaw angle (around 10–25 deg), high wind speeds, and moderate to high negative veer are found to lead to severe SIV. This study should serve as a guiding tool to decide the scope of the more computationally expensive simulations such as high-fidelity CFD-based aeroelastic simulations which can provide a more accurate description of SIV.

Development of seismic demand prediction models for bridges based on probability approach using symbolic regression method

The resilience assessment of existing structures and designing new resilient ones need to have a precise estimation of seismic responses of the structures traditionally carried out by time-consuming calculations of finite element method. This paper presents interpretable, reliable, and fast prediction models developed by two novel branches of the symbolic regression approach (multigene genetic programming and multi biogeography-based programming) in a probability manner for curved bridges. These newly developed techniques seek to find interpretable equations with arbitrary forms that conform to a specific dataset. The reliable prediction models were developed considering the different uncertainties involved, including mechanical, geometrical, structural, and seismic uncertainties. Due to a large number of input variables for the bridge, the evolutionary correlation coefficient was employed to identify the most influential parameters for seismic demands of bridge components. Parameters with the highest correlation were set as inputs for symbolic regression algorithms, generating closed-form mathematical expressions to predict seismic demands. The resulting explicit models present a fast and accurate model that do not require further simulation or the seismic demand model assumptions to develop bridge fragility curves. In comparison to cloud approach and multiple strips analysis, the new method generates fragility curves with less dispersion.

Cascaded Fuzzy Logic for Adaptive Cruise Control

Application of fuzzy logic is a powerful approach that could be applied in a large number of disciplines, starting with engineering control systems, as shown here, but also in other business areas. After a short introduction to fuzzy logic, its application for adaptive cruise control (ACC) is presented. ACC is a driver assistance feature that deals with the problem of speed control, while keeping the safe distance from the vehicle ahead. In the hierarchy of autonomous vehicles autonomy levels, as defined by Society of Automotive Engineers (SAE) International, adaptive cruise control appears in the vehicles at the level 1 and above. We developed a fuzzy logic controller where controlled variables are speed and distance. Input variables include weather conditions, style or mode of driving, vehicle speed and steering angle. A large number of input variables improve control but lead to a large fuzzy rules table. Because of that, in the design presented here, a tree of connected fuzzy inference systems (FIS) is applied. Fuzzy inference systems with a smaller number of variables are developed, algorithms explained, rule base defined, and obtained control surfaces presented. This approach requires less processing time enabling real time applications. Since the rules are defined based on drivers’ experiences, fuzzy logic control systems make decisions in the same way as humans do, i.e., as experience drivers. This paper gives a comprehensive presentation of a novel cascaded fuzzy system development. This novel design also involves algebraic subtraction performed through a FIS subsystem.

Physics-inspired architecture for neural network modeling of forces and torques in particle-laden flows

We present a physics-inspired neural network (PINN) model for direct prediction of hydrodynamic forces and torques experienced by individual particles in stationary arrays of randomly distributed spheres. In line with our findings, it has recently been demonstrated that conventional fully connected neural networks (FCNN) are incapable of making accurate predictions of force variations in a stationary random array of spheres. The problem arises due to the large number of input variables (i.e., the locations of individual neighboring particles) leading to an overwhelmingly large number of training parameters in a fully connected architecture. Given the typically limited size of training datasets that can be generated by particle-resolved simulations, the NN becomes prone to missing true patterns in the data, ultimately leading to overfitting. Inspired by our observations in developing the microstructure-informed probability-driven point-particle (MPP) model in a previous work, we incorporate two main features in the architecture of the present PINN model: (1) superposition of pairwise hydrodynamic interactions between particles, and (2) sharing training parameters between NN blocks that model neighbor influences. These strategies helps to substantially reduce the number of free parameters and thereby control the model complexity without compromising accuracy. We demonstrate that direct force and torque prediction using NNs is indeed possible, provided that the model structure corresponds to the underlying physics of the problem. For a Reynolds number range of 2⩽Re⩽150 and solid volume fractions of 0.1⩽ϕ⩽0.4, the PINN’s predictions prove to be as accurate as those of other currently available microstructure-informed models.

Tourism demand forecasting using stacking ensemble model with adaptive fuzzy combiner

Over the last decades, several soft computing techniques have been applied to tourism demand forecasting. Among these techniques, a neuro-fuzzy model of ANFIS (adaptive neuro-fuzzy inference system) has started to emerge. A conventional ANFIS model cannot deal with the large dimension of a dataset, and cannot work with our dataset, which is composed of a 62 time-series, as well. This study attempts to develop an ensemble model by incorporating neural networks with ANFIS to deal with a large number of input variables for multivariate forecasting. Our proposed approach is a collaboration of two base learners, which are types of the neural network models and a meta-learner of ANFIS in the framework of the stacking ensemble. The results show that the stacking ensemble of ANFIS (meta-learner) and ANN models (base learners) outperforms its stand-alone counterparts of base learners. Numerical results indicate that the proposed ensemble model achieved a MAPE of 7.26% compared to its single-instance ANN models with MAPEs of 8.50 and 9.18%, respectively. Finally, this study which is a novel application of the ensemble systems in the context of tourism demand forecasting has shown better results compared to those of the single expert systems based on the artificial neural networks.

Application of the BP Neural Network Model of Gray Relational Analysis in Economic Management

According to the BP neural network algorithm, for the problem that the input variables of the BP network cannot be automatically optimized during the modeling process of the multivariable complex system, it is used to establish a neural network optimization algorithm according to the gray correlation analysis (GM2 BPANN). Using the data of China’s grain production forecast, the stepwise regression method and the gray GM (1, N) model method were compared and tested. The results show that the new model can comprehensively and extensively process a large number of input variables by using the concept of the gray correlation degree, without having to go through special subjective screening and hence improving the adaptability of the BP network, and at the same time, it can obtain better prediction accuracy and stability. Through the empirical test, the prediction ability of the three methods of regression, GM (1, N) gray system, and GM‐BPANN model is compared. It is proved that the GM‐BPANN optimization algorithm that combines the gray relational analysis and BP neural network method can enhance the multivariable processing ability and network adaptability of the BP network and has good prediction accuracy and stability.

Aberration analysis and compensate method of fully connected neural network in deep ultraviolet lithography

Thermal aberration of a projection lens is caused by lens heating and leads to degradation of image quality. Two tasks have been undertaken in this article: first, analysis of the impact of single aberration on imaging quality, and second, use of machine learning to calculate the combination of Zernike aberration coefficients to reduce the influence of aberration on the process window. The impact of aberration on deep ultraviolet (DUV) lithography imaging has been analyzed separately for 16 Zernike terms. The influence of single Zernike coefficient on depth of focus, mask error enhancement factor, and image log slope is comprehensively analyzed by source mask optimization. However, more than one Zernike term is present in the aberration of a real DUV optics. A fully connected neural network (FCNN) model based on machine learning is widely used in the screening of multivariate experiments and can solve the problem of the large number of input variables and the random values of the input data. Its network structure is simple to debug, with fast screening speed and high precision. To make reasonable budget for aberration, an FCNN model is used to find some combinations of Zernike coefficients for different lithographic performance indicators. This implies that the FCNN model has the ability to select a better Zernike combination while each individual Zernike term value falls in the range of −30 to 30 mλ, which plays a guiding role in DUV optics design.

Modeling biogas production from anaerobic wastewater treatment plants using radial basis function networks and differential evolution

This study presents a new method for modeling biogas production obtained from anaerobic digestion treatment plants with increased accuracy. The method is based on artificial neural networks (ANNs) and more specifically on the efficient architecture of radial basis function (RBF) networks. A novel RBF training scheme is proposed, based on the non-symmetric fuzzy means (NSFM) algorithm, which has been shown to offer increased accuracy compared to other ANN methods, but cannot handle efficiently a large number of input variables. As this is the case in biogas production modeling, the algorithm is enhanced with an optimizer based on differential evolution (DE), which helps to properly tune the algorithm, ultimately boosting the accuracy of the produced models. The proposed approach is applied for modeling the biogas production on a real world, full-scale operational wastewater treatment plant. A comparison study shows the superiority of the proposed model, compared to different machine learning approaches.

Pavement Surface Condition Index Prediction Based on Random Forest Algorithm

With the rise of three-dimensional road detection technology, more and more roads have begun to use three-dimensional road detection vehicles to detect road conditions. The pavement surface condition index (PCI), an important indicator of pavement performance, was used as the research target, and the PCI used the actual pavement data collected by ARAN9000 three-dimensional, multifunctional road detection vehicle. First, the data mining technology is used to consider the factors such as pavement distress, environment, and pavement structure, and to process and analyze the data related to the pavement surface condition index of a certain highway in Ontario, Canada, such as data cleaning and feature screening. Then, a machine learning prediction model of a pavement surface condition index was constructed, and the complex correlation coefficients (R2) of the multiple linear regression model, the neural network model, and the random forest model were 0.562, 0.711, and 0.895. Compared with the neural network model, the accuracy of the random forest model in predicting the pavement surface condition index was improved by 0.184, the error was reduced by 1.599, and the training speed was improved by 33 s. Finally, the random forest model with high precision is selected for optimization. Due to the large number of input variables, it is impossible to determine which line of data is the outlier through simple statistical analysis, so after establishing and predicting the model, the outlier is determined and deleted through the fitting effect between the predicted value and the real value. Then the modified data was used to retrain the model to achieve the optimal training of the current model. The results show that the prediction efficiency and accuracy of the improved random forest model are higher, and the R2 reaches 0.898. The proposed pavement condition index prediction model is accurate and effective, and the prediction results can assist the maintenance department to make more scientific and reasonable maintenance decisions.

Multi-objective constrained optimization for energy applications via tree ensembles

Energy systems optimization problems are complex due to strongly non-linear system behavior and multiple competing objectives, e.g. economic gain vs. environmental impact. Moreover, a large number of input variables and different variable types, e.g. continuous and categorical, are challenges commonly present in real-world applications. In some cases, proposed optimal solutions need to obey explicit input constraints related to physical properties or safety-critical operating conditions. This paper proposes a novel data-driven strategy using tree ensembles for constrained multi-objective optimization of black-box problems with heterogeneous variable spaces for which underlying system dynamics are either too complex to model or unknown. In an extensive case study comprised of synthetic benchmarks and relevant energy applications we demonstrate the competitive performance and sampling efficiency of the proposed algorithm compared to other state-of-the-art tools, making it a useful all-in-one solution for real-world applications with limited evaluation budgets.

Application of Random Forest Algorithm for the Prediction of Online Food Delivery Service Delay

Online shopping industry nowadays has been growing rapidly with the evolution of technology. Consumers have started to shop according to certain criteria with the spread of the online shopping sector. One of the sectors that enlightens the future customer in terms of service quality by getting feedback from purchases (comments or ratings) is the online food sector. In this study, a classification study is conducted to investigate the observance with the fast delivery criteria, which is one of the cornerstone criteria in the online food industry. Random Forest (RF) algorithm is applied for the classification. The most important advantage of the RF is it handles a large number of input variables also it is speediness. More than that RF algorithm reduces the overfitting problem and as a result variance is small and therefore it improves the accuracy. The application is implemented through R programming language. In this study, online food delivery variable is created as two categories (On time or Early and Late) and is estimated by RF algorithm that is applied this data for the first time. According to the results, the correct classification rate of the testing data for the estimation of the online food delivery status variable is found as 95.85%. In addition, the performances of the restaurants are compared for the customers. It turns out that the traffic situation does not greatly affect the result of the delivery status. As a result, RF algorithm is applied to the data obtained by web scraping techniques and the delivery status performance of restaurants is revealed with this study.

Implementation and Performance Evaluation of a Bivariate Cut-HDMR Metamodel for Semiconductor Packaging Design Problems with a Large Number of Input Variables

A metamodeling technique based on Bivariate Cut High Dimensional Model Representation (Bivariate Cut HDMR) is implemented for a semiconductor packaging design problem with 10 design variables. Bivariate Cut-HDMR constructs a metamodel by considering only up to second-order interactions. The implementation uses three uniformly distributed sample points (s = 3) with quadratic spline interpolation to construct the component functions of Bivariate Cut-HDMR, which can be used to make a direct comparison with a metamodel based on Central Composite Design (CCD). The performance of Bivariate Cut-HDMR is evaluated by two well-known error metrics: R-squared and Relative Average Absolute Error (RAAE). The results are compared with the performance of CCD. Bivariate Cut HDMR does not compromise the accuracy compared to CCD, although the former uses only one-fifth of sample points (201 sample points) required by the latter (1045 sample points). The sampling schemes and the predictions of cut-planes and boundary-planes are discussed to explain possible reasons for the outstanding performance of Bivariate Cut HDMR.

Multiple doubling: a simple effective construction technique for optimal two-level experimental designs

Design of experiment is an efficient statistical methodology of establishing which input variables are important (have significant effects) in an experiment (process) and the conditions under which these inputs should work to optimize the outputs of that process. Two-level designs are widely used in high-tech industries and manufacturing for productivity and quality improvement experiments. The construction of (nearly) optimal two-level designs for real-life experiments with large number of input variables can be quite challenging. The practice demonstrated that the existing techniques are complex, highly time-consuming, produce limited types of designs, and likely to fail in large experiments (i.e., optimal results are not expected). To overcome these significant problems, this article gives a simple and effective technique for constructing large two-level designs with good statistical properties. To meet practical needs in different fields, the statistical properties of the generated designs by the new technique are investigated from four basic perspectives: minimizing the similarity among the experimental runs, minimizing the aliasing among the input variables, maximizing the resolution, and filling the experimental domain as uniformly as possible. New recommended saturated orthogonal main effect plans and uniform orthogonal arrays of strength three with thousands or even millions of runs and factors are generated via the new technique without recourse to optimization software.

FUZZY UNORDERED RULE USING GREEDY HILL CLIMBING FEATURE SELECTION METHOD: AN APPLICATION TO DIABETES CLASSIFICATION

Diabetes classification is one of the most crucial applications of healthcare diagnosis. Even though various studies have been conducted in this application, the classification problem remains challenging. Fuzzy logic techniques have recently obtained impressive achievements in different application domains especially medical diagnosis. Fuzzy logic technique is not able to deal with data of a large number of input variables in constructing a classification model. In this research, a fuzzy logic technique using greedy hill climbing feature selection methods was proposed for the classification of diabetes. A dataset of 520 patients from the Hospital of Sylhet in Bangladesh was used to train and evaluate the proposed classifier. Six classification criteria were considered to authenticate the results of the proposed classifier. Comparative analysis proved the effectiveness of the proposed classifier against Naive Bayes, support vector machine, K-nearest neighbour, decision tree, and multilayer perceptron neural network classifiers. Results of the proposed classifier demonstrated the potential of fuzzy logic in analyzing diabetes patterns in all classification criteria.