Grid Search Algorithm Research Articles

Integrative machine learning model for predicting patient-oriented composite endpoints in acute myocardial infarction: a neuroendocrine biomarker approach

Abstract Background Acute Myocardial Infarction (AMI) represents a significant clinical challenge with diverse outcomes. Predicting patient-oriented composite endpoints (POCE) (all-cause death, any stroke, any AMI, or any revascularization) can significantly enhance post-AMI care. Incorporating clinical data with neuroendocrine biomarkers may offer improved prognostic capabilities. Objective To develop and validate a machine learning model to predict POCE in AMI patients, utilizing clinical parameters and neuroendocrine biomarkers. Methods There was prospective, observational, single-center study. Three approaches have been used to obtain the predictions for POCE event. First, as the dataset was imbalanced, an adaptive synthetic (ADASYN) sampling approach was used to generate synthetic instances, particularly focusing on those that are difficult-to-learn by using a weighted distribution. Then, random forest classifier (RF) was used to build a machine learning model to predict POCE. The model was tuned via a grid-search algorithm for optimal hyperparameters and validated using a 10-fold stratified cross-validation. Finally, the feature importance was determined by Shapley Additives, which measure the average marginal contribution of a feature value across all potential feature combinations. For the comparative purpose, the Gini index, which also shows the feature importance but in terms of mean decrease in impurity, was calculated. Results The study incorporated data from 315 patients, examining 47 variables and identifying 72 instances of POCE. After applying the ADASYN algorithm, the class distribution within the dataset was effectively equalized, facilitating the training of a robust RF model. Upon training with 252 instances, the model distinguished POCE with an accuracy of 83.8%, demonstrating a sensitivity of 80% and a specificity of 86%. During 10-fold cross-validation, the model's accuracy slightly dipped to 75%, with a sensitivity of 56% and specificity of 82%, indicating a balance between generalizability and overfitting. Testing mirrored these results, underscoring the model’s consistent performance. Notably, SHAP value analysis highlighted C-Reactive Protein (CRP) and alanine transaminase (ALT) as the most influential features, underscoring their significance in the context of AMI (Figure 1-2). The predictive power of each feature was further demonstrated by the mean decrease in Gini index (Figure 1-2). Conclusion Our study highlights the feasibility of employing machine learning to predict POCE post-AMI using an integrative dataset of clinical and neuroendocrine markers. The predictive model has the potential to revolutionize post-AMI care by allowing clinicians to identify high-risk patients early, tailor interventions, and allocate resources efficiently. Future research could explore the integration of this model into clinical workflows and its impact on patient outcomes.

Precise grading of non-muscle invasive bladder cancer with multi-scale pyramidal CNN

The grading of non-muscle invasive bladder cancer (NMIBC) continues to face challenges due to subjective interpretations, which affect the assessment of its severity. To address this challenge, we are developing an innovative artificial intelligence (AI) system aimed at objectively grading NMIBC. This system uses a novel convolutional neural network (CNN) architecture called the multi-scale pyramidal pretrained CNN to analyze both local and global pathology markers extracted from digital pathology images. The proposed CNN structure takes as input three levels of patches, ranging from small patches (e.g., 128×128\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$128 \ imes 128$$\\end{document}) to the largest size patches (512×512\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$512 \ imes 512$$\\end{document}). These levels are then fused by random forest (RF) to estimate the severity grade of NMIBC. The optimal patch sizes and other model hyperparameters are determined using a grid search algorithm. For each patch size, the proposed system has been trained on 32K patches (comprising 16K low-grade and 16K high-grade samples) and subsequently tested on 8K patches (consisting of 4K low-grade and 4K high-grade samples), all annotated by two pathologists. Incorporating light and efficient processing, defining new benchmarks in the application of AI to histopathology, the ShuffleNet-based AI system achieved notable metrics on the testing data, including 94.25% ± 0.70% accuracy, 94.47% ± 0.93% sensitivity, 94.03% ± 0.95% specificity, and a 94.29% ± 0.70% F1-score. These results highlight its superior performance over traditional models like ResNet-18. The proposed system’s robustness in accurately grading pathology demonstrates its potential as an advanced AI tool for diagnosing human diseases in the domain of digital pathology.

Extraction method for characterizing microcracks on Si3N4 ceramic bearings rolling element with contour segmentation

AbstractAiming at the fuzzy diffusion characteristics of microcrack regions in Si3N4 ceramic bearing rolling elements, an adaptive region segmentation method is proposed to achieve precise extraction of microcrack features. The fuzzy diffusion effect of the microcrack region is analyzed, and a structural complexity matrix is defined to represent the degree of structural variation in the image. A threshold point is determined through histogram analysis, and the matrix dispersion rate is calculated using a box plot to update the correction factor, optimizing the overlap of noise in the microcrack image. The causes of distortion in the microcrack region are analyzed, and pixel points are assigned based on spatial metrics and LAB color features. The fuzzy boundaries and overlapping regions are segmented using the fuzzy C‐means clustering algorithm. A grid search algorithm is embedded to quantitatively evaluate the optimal combination of hyperparameters, enabling comprehensive extraction of microcrack features in Si3N4 ceramic bearing rolling elements. The optimized image achieves an average peak signal‐to‐noise ratio of 39.3, an information fidelity criterion of 7.9328, and an average extraction accuracy of approximately 0.92, effectively overcoming the impact of the fuzzy diffusion effect on the accuracy of microcrack feature extraction in Si3N4 ceramic bearing rolling elements.

Enhancing SRM Performance through Multi-Platform Optimization and FPGA-Based Real-Time Simulation

This paper proposes a multi-platform algorithm methodology in order to define firing angles of torque sharing functions (TSFs) for the indirect torque control of switched reluctance motors (SRM) through a hardware-in-the-loop (HIL) system. This proposal is used to achieve optimal levels of torque ripple and losses with accuracy and reliability, while taking advantage of real-time simulation. The analysis is performed assuming steady-state conditions of speed and torque reference for levels below the base speed, aiming to obtain firing angles ensuring optimal tracking performance. A novel methodology is proposed by using a grid search algorithm that handles the communication between Python, Code Composer, and the Typhoon HIL device. For that, an experimental data FPGA-based model with 500~ns of simulation time step is used to ensure highly accurate dynamic responses of current and electromagnetic torque. Moreover, controller-HIL (C-HIL) is used to have a safe and high fidelity testing environment, allowing rapid testing before transitioning to an experimental test bench. The simulation results are experimentally validated, demonstrating that the proposed strategy is effective and ensures optimal performance, taking into account peripheral systems of A/D, signal conditioning, PWM, and sensor emulation.

Leveraging Deep Learning and Generative AI for Predicting Rheological Properties and Material Compositions of 3D Printed Polyacrylamide Hydrogels.

Artificial intelligence (AI) has the ability to predict rheological properties and constituent composition of 3D-printed materials with appropriately trained models. However, these models are not currently available for use. In this work, we trained deep learning (DL) models to (1) predict the rheological properties, such as the storage (G') and loss (G") moduli, of 3D-printed polyacrylamide (PAA) substrates, and (2) predict the composition of materials and associated 3D printing parameters for a desired pair of G' and G". We employed a multilayer perceptron (MLP) and successfully predicted G' and G" from seven gel constituent parameters in a multivariate regression process. We used a grid-search algorithm along with 10-fold cross validation to tune the hyperparameters of the MLP, and found the R2 value to be 0.89. Next, we adopted two generative DL models named variational autoencoder (VAE) and conditional variational autoencoder (CVAE) to learn data patterns and generate constituent compositions. With these generative models, we produced synthetic data with the same statistical distribution as the real data of actual hydrogel fabrication, which was then validated using Student's t-test and an autoencoder (AE) anomaly detector. We found that none of the seven generated gel constituents were significantly different from the real data. Our trained DL models were successful in mapping the input-output relationship for the 3D-printed hydrogel substrates, which can predict multiple variables from a handful of input variables and vice versa.

Biometric CNN Model for Verification Based on Blockchain and Hyperparameter Optimization

Nowadays, fingerprints as biometrics are among the most popular means of identity verification for various applications. However, they are susceptible to theft, tampering, or alteration by attackers after storage. Hence, it is critical to guarantee the privacy of these fingerprint templates because standard privacy techniques are not secure enough. Additionally, fingerprint templates are verified using a deep learning model to distinguish between authentic and fake fingerprints, making them more protected and secure by storing them inside a blockchain, which has become the most common secure technique in recent years. This paper implements the proposed efficient and secure biometric system for verification based on blockchain technology and hyperparameter optimization. First, for the storing phase, each user’s authentic fingerprint template, private, and public keys are saved in the block and linked to the previous block in the chain by a hash function. If a hacker attempts to assault a fingerprint, all prior blocks must be changed. Second, for the authentication phase, the user logs in with his fingerprint and looks up the required template in the chain. If the required fingerprint exists, it creates a new block with the login details, and the number 1 is returned, which means that the authentication is valid; if it does not exist, it is verified to see if it is authentic or fake using the proposed biometric convolutional neural network (CNN). The proposed CNN uses the Grid Search (GS) algorithm to tune hyperparameters to distinguish between authentic and fake fingerprints. The SOCOFing dataset is used for evaluating our experiment. According to the experimental results, the proposed CNN model achieved the highest accuracy of 99.52%. As a result of the blockchain, our system can return authentication information after looking up the chain in 300 ms.

Analyzing and Optimizing the Distribution of Blood Lead Level Testing for Children in New York City: A Data-Driven Approach.

This study investigates blood lead level (BLL) rates and testing among children under 6 years of age across the 42 neighborhoods in New York City from 2005 to 2021. Despite a citywide general decline in BLL rates, disparities at the neighborhood level persist and are not addressed in the official reports, highlighting the need for this comprehensive analysis. In this paper, we analyze the current BLL testing distribution and cluster the neighborhoods using a k-medoids clustering algorithm. We propose an optimized approach that improves resource allocation efficiency by accounting for case incidences and neighborhood risk profiles using a grid search algorithm. Our findings demonstrate statistically significant improvements in case detection and enhanced fairness by focusing on under-served and high-risk groups. Additionally, we propose actionable recommendations to raise awareness among parents, including outreach at local daycare centers and kindergartens, among other venues.

Study on detection of pesticide residues in tobacco based on hyperspectral imaging technology.

Tobacco is a critical economic crop, yet its cultivation heavily relies on chemical pesticides, posing health risks to consumers, therefore, monitoring pesticide residues in tobacco is conducive to ensuring food safety. However, most current research on pesticide residue detection in tobacco relies on traditional chemical methods, which cannot meet the requirements for real-time and rapid detection. This study introduces an advanced method that combines hyperspectral imaging (HSI) technology with machine learning algorithms. Firstly, a hyperspectral imager was used to obtain spectral data of tobacco samples, and a variety of spectral pre-processing technologies such as mean centralization (MC), trend correction (TC), and wavelet transform (WT), as well as feature extraction methods such as competitive adaptive reweighted sampling (CARS) and least angle regression (LAR) were used to process the spectral data, and then, grid search algorithm (GSA) is used to optimize the support sector machine (SVM). The optimized MC-LAR-SVM model achieved a pesticide classification accuracy of 84.1%, which was 9.5% higher than the original data model. The accuracy of the WT-TC-CARS-GSA-SVM model in the fenvalerate concentration classification experiment was as high as 91.8 %, and it also had excellent performance in other metrics. Compared with the model based on the original data, the accuracy, precision, recall, and F1-score are improved by 8.3 %, 8.2 %, 7.5 %, and 0.08, respectively. The results show that combining spectral preprocessing and feature extraction algorithms with machine learning models can significantly enhance the performance of pesticide residue detection models and provide robust, efficient, and accurate solutions for food safety monitoring. This study provides a new technical means for the detection of pesticide residues in tobacco, which is of great significance for improving the efficiency and accuracy of food safety detection.

Classification of Unisba Students' Graduation Time using Support Vector Machine Optimized with Grid Search Algorithm

Support Vector Machine is a classification method that finds the optimal hyperplane to separate two data classes. SVM has much better generalization performance than other methods. However, SVM needs to improve in determining hyperparameter values. Therefore, parameter optimization is necessary to determine the optimal hyperparameter value. Grid search is one of the parameter optimization methods that can improve the quality of SVM models. This study aims to assess the level of accuracy in predicting student graduation times by using five features that affect it. This study shows that the resulting SVM model optimized with the Grid Search Algorithm is quite consistent and prevents overfitting. By utilizing the results of SVM modelling, UNISBA is expected to improve the quality of graduates. The risk of delays in graduation can be considered early by paying attention to the background and achievements of students

Application of LSTM neural network in the research of cyclical industries: taking the automotive industry as an illustrative case

Anticipating future developmental trends within an industry stands as one of the paramount responsibilities for industry researchers. Nevertheless, owing to the intricate and subjective nature inherent in industry research, coupled with researchers usually having to rely on their own subjective judgments when making predictions, the forecasting outcomes frequently prove unsatisfactory. In order to address this issue, this study introduces deep learning technology into the industry research field and employs Long & Short-Term Memory (LSTM) networks to forecast a typical cyclical industry - the automotive industry. In the specific experiment, Adam optimization algorithm is employed to enhance model training speed, followed by utilization of the grid search algorithm for optimal parameter selection. Subsequently, the obtained optimal model is utilized for LSTM multivariate sequence prediction in order to forecast the future development trend of the automotive industry over the next six months. The ultimate outcome demonstrates that the LSTM model accurately anticipated the forthcoming U-shaped rebound trend in the automotive industry, albeit with some limitations in predicting precise numerical values and timing. Nevertheless, it aligns closely with the actual developmental trajectory. This investigation represents a constructive application of deep learning technology in industrial research, and its final findings hold significant implications for future industry research endeavors.

DETECTION OF FAKE NEWS ON TWITTER USING MACHINE LEARNING: AN XGBOOST-BASED APPROACH WITH SENTIMENT AND SOURCE CHARACTERISTIC ANALYSIS

The spread of fake news on social media platforms is becoming an increasingly alarming problem with fake news becoming more deceptive and harder to detect. Twitter, in particular, poses a significant threat as fake news spreads faster than real news on the platform, enhancing misinformation and leading to serious consequences.This project presents a novel machine learning-based approach for detecting fake news tweets on Twitter using the TruthSeeker 2023 dataset from the University of New Brunswick. As the largest ground truth dataset for fake news detection on social media, it contains over 130,000 crowdsourced tweets, enabling the creation of a broader and more applicable model for real-world scenarios. The algorithm employed in this study leverages the properties of gradient-boosted decision tree models (XGBoost) to develop a novel method for classifying fake and real news tweets. The proposed model preprocesses the data by extracting additional features for each tweet, such as detailed sentiment analysis of both the tweet and the related news statement, as well as features pertaining to the author. These features are added to the tweets feature vector. The enhanced feature vectors are then fed into an XGBoost model with tuned hyperparameters determined through a grid search algorithm to perform binary classification. The additional extracted features increase the robustness of the model by highlighting key differentiating factors between real and fake tweets. The results of this study demonstrate the effectiveness of the proposed algorithm, achieving an accuracy of 0.9335 on over 13,000 unseen tweets.

A Hybrid Grey System Model Based on Stacked Long Short-Term Memory Layers and Its Application in Energy Consumption Forecasting

Accurate energy consumption prediction is crucial for addressing energy scheduling problems. Traditional machine learning models often struggle with small-scale datasets and nonlinear data patterns. To address these challenges, this paper proposes a hybrid grey model based on stacked LSTM layers. This approach leverages neural network structures to enhance feature learning and harnesses the strengths of grey models in handling small-scale data. The model is trained using the Adam algorithm with parameter optimization facilitated by the grid search algorithm. We use the latest annual data on coal, electricity, and gasoline consumption in Henan Province as the application background. The model’s performance is evaluated against nine machine learning models and fifteen grey models based on four performance metrics. Our results show that the proposed model achieves the smallest prediction errors across all four metrics (RMSE, MAE, MAPE, TIC, U1, U2) compared with other 15 grey system models and 9 machine learning models during the testing phase, indicating higher prediction accuracy and stronger generalization performance. Additionally, the study investigates the impact of different LSTM layers on the model’s prediction performance, concluding that while increasing the number of layers initially improves prediction performance, too many layers lead to overfitting.

Research on non-destructive identification technology of rice varieties based on HSI and GBDT

Accurate identification of rice varieties is of great significance for rice planting, field management and storage, and is also a key link in the process of agricultural breeding. In this study, a gradient boosting decision tree (GBDT) model was established based on hyperspectral imaging (HSI) to realize high-speed and non-destructive variety identification of six rice varieties. In this study, the near-infrared hyperspectral images of 600 rice samples of 6 varieties were taken as the research object, and the characteristic spectra of sensitive regions of the sample spectral images were processed by multiplicative scatter correction (MSC), and after the characteristic wavelengths were determined by the importance scores, the GBDT model to realize the identification of rice sample varieties, and the grid search algorithm was used to optimize the four internal parameters of GBDT. The results showed that the established GBDT model for the accuracy of rice variety identification of vitro test set samples reached 95%, indicating that HSI can be used to quickly and non-destructively identify rice varieties, and provide a new idea for batch online non-destructive testing of rice seeds.

$$D_MD_RDF$$: diabetes mellitus and retinopathy detection framework using artificial intelligence and feature selection

AbstractDiabetes mellitus is one of the most common diseases affecting patients of different ages. Diabetes can be controlled if diagnosed as early as possible. One of the serious complications of diabetes affecting the retina is diabetic retinopathy. If not diagnosed early, it can lead to blindness. Our purpose is to propose a novel framework, named $$D_MD_RDF$$ D M D R D F , for early and accurate diagnosis of diabetes and diabetic retinopathy. The framework consists of two phases, one for diabetes mellitus detection (DMD) and the other for diabetic retinopathy detection (DRD). The novelty of DMD phase is concerned in two contributions. Firstly, a novel feature selection approach called Advanced Aquila Optimizer Feature Selection ($$A^2OFS$$ A 2 O F S ) is introduced to choose the most promising features for diagnosing diabetes. This approach extracts the required features from the results of laboratory tests while ignoring the useless features. Secondly, a novel classification approach (CA) using five modified machine learning (ML) algorithms is used. This modification of the ML algorithms is proposed to automatically select the parameters of these algorithms using Grid Search (GS) algorithm. The novelty of DRD phase lies in the modification of 7 CNNs using Aquila Optimizer for the classification of diabetic retinopathy. The reported results concerning the DMD datasets shows that AO reports best performance metrics in the feature selection process with the help of modified ML classifiers. The best achieved accuracy is 98.65% with the GS-ERTC model and max-absolute scaling on the “Early Stage Diabetes Risk Prediction Dataset” dataset. Also, from the reported results concerning the DRD datasets, the AOMobileNet is considered a suitable model for this problem as it outperforms the other modified CNN models with accuracy of 95.80% on the “The SUSTech-SYSU dataset” dataset.

Forecast of chlorophyll-a concentration as an indicator of phytoplankton biomass in El Val reservoir by utilizing various machine learning techniques: A case study in Ebro river basin, Spain

The trophic condition of bodies of water, such as oceans, lakes, and reservoirs, can be accurately assessed thanks to the use of chlorophyll-a, or Chl-a, as an indicator of phytoplankton biomass and abundance. In fact, the main molecule in charge of photosynthesis is Chl-a. This work presents a powerful and reliable nonparametric method for predicting the concentration of Chl-a in El Val reservoir using a dataset containing 240,765 samples: the Support Vector Regression (SVR) with different kinds of kernels. This mathematical SVR-relied model was constructed using five years (2018–2022) of water quality variable monitoring (physico-chemical independent variables) in the El Val reservoir (located in the northeast of Spain). For comparison, M5 model trees, a the Multilayer Perceptron (MLP), that is a particular type of artificial neural network (ANN), and multivariate linear regression (MLR) were also used for the same observed data. The Grid Search (GS) algorithm was employed as an optimizer; this approach greatly improves the regression precision by allowing the optimal kernel parameters to be chosen during the SVR training phase. There are two ways to sum up the findings of this investigation. First, it is determined how relevant each input variable is to the Chl-a concentration in the El Val reservoir. Second, this hybrid GS/SVR-relied model with PUK kernel can accurately predict the Chl-a (R2 and r values were 0.8989 and 0.9499, respectively). The model’s agreement with the observed data amply proves the remarkable efficacy of this creative strategy.

Sustainable energy transition optimization through decentralized hybrid energy systems with various energy storage technologies under multi-criterion indices: A mid-career repowering scenarios

Sustainable energy transition (SET) refers to the formulation of effective policies for quantifying the paradigm shift from fossil fuels to carbon-neutral technologies. Despite the anticipation of sustainable solutions, many developing nations continue to experience unsustainable energy regimes. Understanding the potential of this transition demands a multifaceted approach, emphasizing techno-economic optimization and performance enhancement to address future decentralized energy constraints. Although one of several ways to enable SET, the consideration of renewable energy sources (RESs) coupled with energy storage technologies (ESTs), which can offer multi-dimensional decision flexibility in present energy policies, is quite restricted. Based on a comparative analysis of various ESTs, this paper provides insight exploration of the SET framework by leveraging a diverse range of renewable potentials. Utilizing a proprietary derivative-free algorithm (PDFA) and an original grid-search algorithm (OGSA), the study focuses on load forecasting across different time horizons using a real-time dataset from multiple load centers in Pakistan. Load forecasting is carried out based on a combination of seasonal naïve forecasting (SNF), seasonal and trend decomposition using loess forecasting (STLF), and artificial neural networks (ANN). Results reveal a minimum Levelized cost of energy (LCOE) ranging from $0.105/kWh to $0.109/kWh in the presence of solar and dispatchable renewable-based hybrid configurations. For battery ESTs, based on NPC and LCOE, the feasibility is observed in the order of lead-acid (La), lithium-ion (Li-ion), and vanadium redox flow (VRF) technology. From a technical and capacity shortage fraction (CSF) perspective, Li-ion performed better than both La and VRF in terms of storage depletion as well as annual losses. Finally, a mid-career repowering analysis examines the behavior of energy storage technologies (ESTs), while sensitivity analyses and results validation highlight the significance of the proposed study's findings.

Optimzing Android Program Malware Classification Using GridSearchCV Optimized Random Forest

The growing number of smartphones, particularly Android powered ones, has increased public awareness of the security concerns posed by malware and viruses. While machine learning models have been studied for malware prediction in this field, methods for precise identification and classification still require improvement for the perfect detection of malwares and minimizing the cracks on machine learning based classification. Detection accuracy that ranges from 93% to 95% has been observed in prior research, indicates room for improvement. In order to maximize the hyperparameters, this paper suggests improving the Random Forest method by introducing the grid search algorithm which isn’t present in previous studies. A significant increase in classification accuracy is the main aim of the research. We exhibit an outstanding 99% accuracy rate in detecting malware contaminated programs, demonstrating the significance of our technique. The proposed method can be seen as a huge improvement over existing models, achieving near perfection in detection, in contrast to which typically obtained by previous models with the accuracy rate of 95% max on the same dataset. Our approach achieves such high accuracy and provides a novel remedy for the limits of the Android based platforms, particularly when program processing resources are limited. This study confirms the effectiveness of our improved Random Forest algorithm, points to a paradigm shift in malware detection, and heightened cybersecurity measures for the rapidly growing smartphone market.

Separation of source, attenuation and site parameters of 2 moderate earthquakes in France: an elastic radiative transfer approach

Summary An accurate magnitude estimation is necessary to properly evaluate seismic hazard, especially in low to moderate seismicity areas such as Metropolitan France. However, magnitudes of small earthquakes are subject to large uncertainties caused by major high-frequency propagation effects which are generally not properly considered. To address this issue, we developed a method to separate source, attenuation and site parameters from the elastic radiative transfer modeling of the full energy envelopes of seismograms. The key feature of our approach is the treatment of attenuation -both scattering and absorption- in a simple but realistic velocity model of the Earth’s lithosphere, including a velocity discontinuity at the Moho. To reach this goal, we developed a 2-step inversion procedure, allowing first to extract attenuation parameters for each source-station path from the whole observed energy envelope using the Levenberg-Marquardt and grid-search algorithms, then to determine site amplification and the source displacement spectrum from which the moment magnitude Mw is extracted. In the first step, we use the forward modeling procedure of Heller et al. (2022) in order to simulate energy envelopes by taking into account the full treatment of wave polarization, the focal mechanism of the source and the scattering anisotropy. The inversion procedure is then applied to the 2019 ML 5.2 Le Teil and 2014 ML 4.5 Lourdes earthquakes which both occurred in southern France. Data from 6 stations are selected for each event. The inversion results confirm a significant variability in the attenuation parameters (scattering and absorption) at regional scale and a strong frequency dependence. Scattering appears to be stronger towards the French Alps and Western Pyrenees. Absorption is stronger as frequency increases. Although not very resolvable, the mechanism of scattering appears to be forward or very forward. By inverting the source spectrum, we determine moment magnitudes Mw of 5.02 ± 0.17 for the Le Teil earthquake and 4.17 ± 0.15 for the Lourdes earthquake.

Forecasting U.S. Fossil Energy Consumption: Advancing Accuracy with Multilayer Perceptron and Residual Learning

Accurate midterm forecasts of fossil fuel consumption are essential for effective energy planning, economic management, and resource allocation. While machine learning models have demonstrated their efficacy in handling large-scale nonlinear datasets, many, including Multilayer Perceptrons (MLPs), suffer from performance degradation with increased depth. Fortunately, recent studies have revealed that Residual Networks (ResNets) can mitigate or even overcome this challenge. In this paper, we propose a Weighted Residual Network based on MLP to enhance predictive performance. We employ the Adam algorithm for model training and utilize the Gridsearch algorithm for hyperparameter tuning. In the application section, we develop predictive models using three case studies: natural gas, petroleum, and total fossil fuel consumption. We validate the effectiveness of our proposed model and compare it with ten other machine learning models. Our findings demonstrate that our proposed model consistently outperforms others in all three cases, underscoring its superior performance in midterm forecasting of fossil fuel consumption.

VOC data-driven evaluation of vehicle cabin odor: from ANN to CNN-BiLSTM.

Emissions of volatile organic compounds (VOCs) in vehicles represent a significant problem, causing unpleasant odors. To mitigate VOCs and odors in vehicles, it is critical to choose interior parts with low odor and VOC emissions. However, prevailing odor evaluation methods are subjective, costly, and potentially harmful to the health of evaluators. In this study, we analyzed 139 automotive interior parts and 92 vehicles, establishing a cost-effective, data-driven method for odor evaluation. The contents of benzene, toluene, ethylbenzene, xylene, styrene, formaldehyde, acetaldehyde, acrolein, and total volatile organic compounds (TVOC) were detected by thermal desorption gas chromatography-mass spectrometry (TD-GC/MS) and high-performance liquid chromatography with an ultraviolet detector (HPLC-UV). Professional odor evaluators assessed the odors, identifying intensity levels from 2.0 to 4.5 in interior parts and 2.5 to 3.5 in whole vehicles. Leveraging this data, we applied four supervised learning algorithms to develop predictive models for the odor intensity of both interior parts and entire vehicles. During model training, we implemented early stopping techniques for the artificial neural network (ANN) and convolutional neural network-bidirectional long short-term memory (CNN-BiLSTM) models, while optimizing the support vector machine (SVM) and extreme gradient boosting (XGBoost) models using the GridSearch algorithm. The evaluation results reveal that the CNN-BiLSTM model performs the best, achieving an average accuracy of 89% for unknown samples within an odor intensity level of 0.5. The root mean square error (RMSE) is 0.24, and the mean absolute error (MAE) is 0.08. The model also underwent a sevenfold cross-validation, achieving an accuracy of 83.43%. Additionally, we employed SHapley Additive exPlanations (SHAP) for the interpretative analysis of the model, which confirmed the consistency of each VOC's odor contribution with human olfactory rules. By predicting odors based on VOCs through supervised learning, this study reduces the costs and enhances the efficiency and applicability of odor assessment across various vehicle interiors.