Performance Of Hybrid Model Research Articles

Exploring the performance and interpretability of hybrid hydrologic model coupling physical mechanisms and deep learning

Recently, differentiable modeling techniques have emerged as a promising approach to bidirectionally integrating neural networks and hydrologic models, achieving performance levels close to deep learning models while preserving the ability to output physical states and fluxes. However, there remains a lack of systematic exploration into the performance and physical interpretability of hybrid models that use neural networks to replace the runoff generation and routing processes in regionalized modeling. This research developed 12 regionalized hybrid models based on a differentiable parameter learning (DPL) framework, utilizing the Hydrologiska Byråns Vattenbalansavdelning (HBV) model as the foundational backbone. These hybrid models incorporate neural networks to replace the various physical processes within the runoff generation and routing modules. The publicly available CAMELS dataset is employed to evaluate the performance and interpretability of these hybrid models. The results show that while the median Nash-Sutcliffe efficiency (NSE) and Kling-Gupta efficiency (KGE) coefficients for all hybrid models are lower than those of the purely data-driven regionalized long short-term memory neural network (LSTM) model (median NSE: 0.742, median KGE: 0.762), the best-performing hybrid model (median NSE: 0.731, median KGE: 0.761) approaches the LSTM model and has better physical interpretability. Embedding neural networks does not inherently guarantee improved performance and may, in some cases, even result in reduced performance. The degree of performance enhancement is not significantly correlated with the number of embedded neural networks. Compared to replacing the runoff generation process, substituting the routing process with neural networks yields more substantial performance improvements and enables the learning of different routing patterns based on the catchment’s static attributes. This study underscores the importance of reasonably balancing the location, complexity, and quantity of embedded neural networks to achieve a trade-off between model performance and interpretability in hybrid modeling. These insights contribute to advancing regionalized hybrid modeling development.

Advancing Marine Surveillance: A Hybrid Approach of Physics Infused Neural Network for Enhanced Vessel Tracking Using Automatic Identification System Data

In the domain of maritime surveillance, the continuous tracking and monitoring of vessels are imperative for the early detection of potential threats. The Automatic Identification System (AIS) database, which collects vessel movement data over time, including timestamps and other motion details, plays a crucial role in real-time maritime monitoring. However, it frequently exhibits irregular intervals of data collection and intricate, intersecting trajectories, underscoring the importance of analyzing long-term temporal patterns for effective vessel tracking. While Kalman Filters and other physics-based models have been employed to tackle these issues, their effectiveness is limited by their inability to capture long-term dependence and non-linearity in the historical data. This paper introduces a novel approach that leverages Long Short-Term Memory (LSTM), a type of recurrent neural network, renowned for its proficiency in recognizing patterns over extended periods. Recognizing the strengths and limitations of the LSTM model, we propose a hybrid machine-learning algorithm that integrates LSTM with a physics-based model. This combination harnesses the physical laws governing vessel movements alongside data driven pattern mining, thereby enhancing the predictive accuracy of vessel locations. To assess the performance of standalone and hybrid models, various scenarios with different levels of complexity are generated. Furthermore, to simulate real-world data loss conditions often encountered in maritime tracking, temporal data gaps are randomly introduced into the scenarios. The competing approaches are then evaluated using both with time gap and without time gap conditions. Our results show that, although the LSTM model performs better than the physics-based model, the hybrid model consistently outperforms both standalone models across all scenarios. Furthermore, while data gaps negatively impact the accuracy of all models, the performance reduction is minimal for the physics-infused model. In summary, this study not only demonstrates the potential of combining data-driven and physics-based approaches but also sets a new benchmark for maritime vessel tracking.

Incremental Value of Radiomics Features of Epicardial Adipose Tissue for Detecting the Severity of COVID-19 Infection.

Epicardial adipose tissue (EAT) is known for its pro-inflammatory properties and association with Coronavirus Disease 2019 (COVID-19) severity. However, existing detection methods for COVID-19 severity assessment often lack consideration of organs and tissues other than the lungs, which limits the accuracy and reliability of these predictive models. The retrospective study included data from 515 COVID-19 patients (Cohort 1, n=415; Cohort 2, n=100) from two centers (Shanghai Public Health Center and Brazil Niteroi Hospital) between January 2020 and July 2020. Firstly, a three-stage EAT segmentation method was proposed by combining object detection and segmentation networks. Lung and EAT radiomics features were then extracted, and feature selection was performed. Finally, a hybrid model, based on seven machine learning models, was built for detecting COVID-19 severity. The hybrid model's performance and uncertainty were evaluated in both internal and external validation cohorts. For EAT extraction, the Dice similarity coefficients (DSC) of the two centers were 0.972 (±0.011) and 0.968 (±0.005), respectively. For severity detection, the area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), and integrated discrimination improvement (IDI) of the hybrid model increased by 0.09 (p<0.001), 19.3 % (p<0.05), and 18.0 % (p<0.05) in the internal validation cohort, and by 0.06 (p<0.001), 18.0 % (p<0.05) and 18.0 % (p<0.05) in the external validation cohort, respectively. Uncertainty and radiomics features analysis confirmed the interpretability of increased certainty in case prediction after inclusion of EAT features. This study proposed a novel three-stage EAT extraction method. We demonstrated that adding EAT radiomics features to a COVID-19 severity detection model results in increased accuracy and reduced uncertainty. The value of these features was also confirmed through feature importance ranking and visualization.

Comparative Analysis of Machine Learning Models and Hybrid Ensemble Approach’s for Landslide Prediction

The NH-44 Jammu Srinagar National Highway in India is susceptible to landslides, rock falls and shooting stones due to its geological characteristics and steep slopes. This study aims to compare the performance of various Machine Learning (ML) algorithms and hybrid models in predicting landslides using historical data. Seven optimized ML approaches Support Vector Classifier (SVC), Logistic Regression (LR), Decision Tree (DT), K Nearest Classifier (KNC), Random Forest (RF), GaussianNB (GNB) and AdaBoost Classifier (ABC) are used. Additionally, two Hybrid Ensemble methods, the Voting Hybrid Model (VHM) and Stacking Hybrid Model (SHM), are introduced. The performance of each model is evaluated using accuracy, precision, recall, F1-score and AUC metrics. Results indicate that all the models performed well, with hybrid ensemble models surpassing all individual algorithms. The Stacking Hybrid Model (SHM) excels achieving 99.4% accuracy and 98.5% AUC, outperforming the Voting Hybrid Model (VHM). Hybrid models consistently outperform individual models in accuracy and AUC. These proposed methods exhibit robustness and enhanced results in addressing this issue. This framework can help to predict the landslides with high accuracy which can save the lives through timely evacuations from high-risk areas.

Development and testing of hybrid (PNM–CFD) mathematical model and numerical algorithm for description of fluid flows in three-dimensional digital core models

Numerical simulation of fluid flow in porous and fractured rocks is an important task for many industrial applications. The most common approaches to constructing such models are direct (CFD or lattice Boltzmann) and porous network (PNM) modeling. Each approach has its own advantages and disadvantages. This paper presents a hybrid mathematical PNM–CFD model for describing fluid flows in three-dimensional digital core models, which has the high speed performance of PNM approaches and high accuracy of CFD models. Numerical technique has been developed for describing fluid flows in three-dimensional digital core models using a hybrid PNM–CFD model. The numerical technique links a one-dimensional pore network solver and a three-dimensional CFD solver into a combined model in original way by constructing a single pressure field for the entire computation domain. To validate the model, several tests were performed, including flow in straight channels and numerical simulation of fluid flow in a microfluidic chip. The test results have shown the adequacy of the hybrid model performance. The hybrid model for determining pressure drop in a branched network with three-dimensional chambers has an error of no more than 5 % when compared to experimental data. Similarly, the error in calculating velocity does not exceed 7 % when compared to the full three-dimensional calculation. The hybrid model has shown an almost twofold increase in calculation speed compared to the full three-dimensional model.

Stacking Ensemble Artificial Intelligence Model for Heart Disease Diagnosis

Personalized treatment plans, predictive analytics, and artificial intelligence (AI)-driven diagnostics are becoming more and more popular as a way to improve decision-making, expedite operations, and improve patient care. But there are still a number of substantial barriers to overcome, which includes but not limited to issues with user adoption, trust, prejudice, and fairness brought on by resistance from healthcare providers and a lack of confidence in the system's recommendations. To overcome these challenges and realize the full potential of AI-driven solutions, the system's accuracy and safety through meticulous testing and validation of AI algorithms becomes indispensable. This research provides a hybrid AI model that blends three base models with a Meta model in order to diagnose heart disease effectively. The essence is to revalidate the existing AI diagnostic models for cardiac diseases diagnostics and to tackle the concerns impeding the full utilization of the available AI diagnostic system. The study makes use of each model's advantages by merging these various and complimentary algorithms into a stacking ensemble model to create a diagnostic system that is more potent. Using publicly available heart disease data, the model performs remarkably well; it achieves 89% accuracy, 85% recall (sensitivity), 92% specificity, and 89% precision. This hybrid model's performance and proven efficacy are expected to boost trust in the system's recommendations and encourage broader implementation in clinical practice.

Energy-Efficient Anomaly Detection and Chaoticity in Electric Vehicle Driving Behavior.

Detection of abnormal situations in mobile systems not only provides predictions about risky situations but also has the potential to increase energy efficiency. In this study, two real-world drives of a battery electric vehicle and unsupervised hybrid anomaly detection approaches were developed. The anomaly detection performances of hybrid models created with the combination of Long Short-Term Memory (LSTM)-Autoencoder, the Local Outlier Factor (LOF), and the Mahalanobis distance were evaluated with the silhouette score, Davies-Bouldin index, and Calinski-Harabasz index, and the potential energy recovery rates were also determined. Two driving datasets were evaluated in terms of chaotic aspects using the Lyapunov exponent, Kolmogorov-Sinai entropy, and fractal dimension metrics. The developed hybrid models are superior to the sub-methods in anomaly detection. Hybrid Model-2 had 2.92% more successful results in anomaly detection compared to Hybrid Model-1. In terms of potential energy saving, Hybrid Model-1 provided 31.26% superiority, while Hybrid Model-2 provided 31.48%. It was also observed that there is a close relationship between anomaly and chaoticity. In the literature where cyber security and visual sources dominate in anomaly detection, a strategy was developed that provides energy efficiency-based anomaly detection and chaotic analysis from data obtained without additional sensor data.

Designing an explainable bio-inspired model for suspended sediment load estimation: eXtreme Gradient Boosting coupled with Marine Predators Algorithm

This study aimed to develop an accurate and reliable model for predicting suspended sediment load (SL) in river systems, which is crucial for water resource management and environmental protection. While Xtreme Gradient Boosting (XGB), a powerful ensemble machine learning (ML) model, has been employed in previous studies, the novelty of this research lies in the introduction of a hybrid approach that synergistically combines XGB with the bio-inspired Marine Predators Algorithm (XGB-MPA) to estimate SL in the Yeşilirmak River (Turkey). To this end, streamflow (Q) and sediment concentration (SC) values as well as their lag times (1 to 3 month lag times) were fed as input variables – under 9 scenarios – into ML models. A time series of datasets from March 1973 to December 2011 and January 2012 to March 2023 were used for training and testing of ML models, respectively. The superiority of the proposed model (XGB-MPA) compared to two other hybrid models, including XGB-PSO (Particle Swarm Optimization) and XGB-GWO (Grey Wolf Optimization) was also investigated. According to the results, the simultaneous application of Q and SC lag time values as inputs has led to the best SL estimates by XGB-MPA, with XGB-MPA9 (RMSE = 103.7 ton/day; NSE = 0.96) exhibiting the lowest error rates. In addition, XGB-MPA has performed better than XGB in all scenarios, with the lowest and highest reduction in RMSE being 19.3% (scenario 5) and 97.4% (scenario 1), respectively. When comparing the performance of hybrid models, the proposed XGB-MPA model has performed best with MAE, RMSE and NSE of 40.94, 103.7 and 0.96, respectively, in comparison with 816.02, 1063.74 and −2.94 for XGB-PSO and 693.16, 981.68 and −2.37 for XGB-GWO. Further research can include the use of time series of efficient variables extracted from satellite images (e.g. land cover, river morphology, etc.) as model inputs. Highlights Improvement of SL estimation by coupling MPA with XGB model Using delayed combinations of streamflow and sediment concentration as model inputs Superiority of MPA compare to PSO and GWO in SL estimation Greater variations in SHAP caused by sediment concentration compared to streamflow Further studies required on the effects of hydrological and topographical features

Abstract We093: Comparative Analysis of Theoretical Models and Fusion Frameworks for Early Cardiac Dysfunction Detection Using Multimodal Wearable Data

Background: Wearable devices have emerged as promising tools for continuous monitoring of physiological signals, offering opportunities for early detection of cardiac dysfunction. This study aims to compare various theoretical models and fusion frameworks to assess their effectiveness in early detection of cardiac dysfunction. Hypothesis: The study investigates the performance of signal processing-based, machine learning-based, and hybrid models for early cardiac dysfunction detection using multimodal wearable data. Additionally, it explores the efficacy of feature-level fusion, decision-level fusion, and sensor-level fusion frameworks in integrating information from diverse sensors for enhanced detection accuracy.The primary goal of the study is to identify the most effective theoretical models and fusion frameworks for early detection of cardiac dysfunction using multimodal wearable data. By comparing different approaches, the study aims to provide insights into the strengths and limitations of each method and contribute to the development of reliable wearable-based systems for proactive healthcare interventions. Methods/Approach: Three theoretical models—signal processing-based, machine learning-based, and hybrid—were implemented and evaluated for their performance in detecting early signs of cardiac dysfunction. Additionally, three fusion frameworks—feature-level fusion, decision-level fusion, and sensor-level fusion—were investigated to assess their effectiveness in integrating multimodal wearable data. Results/Data: The signal processing-based model achieved a sensitivity of 85%, specificity of 82%, and accuracy of 83% in detecting early signs of cardiac dysfunction. The machine learning-based model outperformed with a sensitivity of 92%, specificity of 88%, and accuracy of 90% . The hybrid model yielded intermediate results with a sensitivity of 89%, specificity of 85%, and accuracy of 87% . Among the fusion frameworks, sensor-level fusion demonstrated superior performance with a sensitivity of 94%, specificity of 90%, and accuracy of 92%. Conclusion: Machine learning-based approaches showed higher efficacy in early cardiac dysfunction detection compared to signal processing-based methods. Sensor-level fusion exhibited the best performance among data fusion frameworks, emphasising the importance of integrating information from diverse wearable sensors for improved accuracy.

Performance analysis of deep unified model for facial expression recognition using convolution neural network

Facial expression recognition has gathered substantial attention in computer vision applications, with the need for robust and accurate models that can decipher human emotions from facial images. Performance analysis of a novel hybrid model combines the strengths of residual network (ResNet) and dense network (DenseNet) architectures after applying preprocessing for facial expression recognition. The proposed hybrid model capitalizes on the complementary characteristics of ResNet's and DenseNet's densely connected blocks to enhance the model's capacity to extract discriminative features from facial images. This research evaluates the hybrid model performance and conducts a comprehensive benchmark against established facial expression recognition convolution neural network (CNN) models. The analysis encompasses key aspects of model performance, including classification accuracy, and adaptability with the LFW dataset for facial expressions such as Anger, Fear, Happy, Disgust, Sad, Surprise, along Neutral. The research observes that the proposed hybrid model is more accurate and efficient computationally than existing models consistently. This performance analysis eliminates information on the hybrid model's perspective to further facial expression recognition research.

A hybrid model for predicting response to risperidone after first episode of psychosis.

Patient response to antipsychotic drugs varies and may be related to clinical and genetic heterogeneity. This study aimed to determine the performance of clinical, genetic, and hybrid models to predict the response of first episode of psychosis (FEP). patients to the antipsychotic risperidone. We evaluated 141 antipsychotic-naïve FEP patients before and after 10 weeks of risperidone treatment. Patients who had a response rate equal to or higher than 50% on the Positive and Negative Syndrome Scale were considered responders (n = 72; 51%). Analyses were performed using a support vector machine (SVM), k-nearest neighbors (kNN), and random forests (RF). Clinical and genetic (with single-nucleotide variants [SNVs]) models were created separately. Hybrid models (clinical+genetic factors) with and without feature selection were created. Clinical models presented greater balanced accuracy 63.3% (confidence interval [CI] 0.46-0.69) with the SVM algorithm than the genetic models (balanced accuracy: 58.5% [CI 0.41-0.76] - kNN algorithm). The hybrid model, which included duration of untreated psychosis, Clinical Global Impression-Severity scale scores, age, cannabis use, and 406 SNVs, showed the best performance (balanced accuracy: 72.9% [CI 0.62-0.84] - RF algorithm). A hybrid model, including clinical and genetic predictors, can provide enhanced predictions of response to antipsychotic treatment.

Streamflow Prediction at the Intersection of Physics and Machine Learning: A Case Study of Two Mediterranean‐Climate Watersheds

AbstractAccurate streamflow predictions are essential for water resources management. Recent studies have examined the use of hybrid models that integrate machine learning models with process‐based (PB) hydrologic models to improve streamflow predictions. Yet, there are many open questions regarding optimal hybrid model construction, especially in Mediterranean‐climate watersheds that experience pronounced wet and dry seasons. In this study, we performed model benchmarking to (a) compare hybrid model performance to PB and machine learning models and (b) examine the sensitivity of hybrid model performance to PB model parameter calibration, structural complexity, and variable selection. Hybrid models were generated by post‐processing process‐based models using Long Short‐Term Memory neural networks. Models were benchmarked within two northern California watersheds that are managed for both municipal water supplies and aquatic habitat. Though model performance varied substantially by watershed and error metric, calibrated hybrid models frequently outperformed both the machine learning model (for 72% of watershed‐model‐metric combinations) and the calibrated process‐based models (for 79% of combinations). Furthermore, hybrid models were relatively insensitive to PB model calibration and structural complexity, but sensitive to PB model variable selection. Our results demonstrate that hybrid models can improve streamflow prediction in Mediterranean‐climate watersheds. Additionally, hybrid model insensitivity to PB model parameter calibration and structural complexity suggests that uncalibrated or less complex PB models could be used in hybrid models without any loss of streamflow prediction accuracy, improving model construction efficiency. Moreover, hybrid model sensitivity to the selection of PB model variables suggests a strategy for diagnosing poorly performing PB model components.

Comparative performance of hybrid model based on discrete wavelet transform and ARIMA models in prediction incidence of COVID-19

ObjectivePublic health surveillance is an important aspect of outbreak early warning based on prediction models. The present study compares a hybrid model based on discrete wavelet transform (DWT) and ARIMA (Autoregressive Integrated Moving Average) for predicting incidence cases due to COVID-19. MethodsIn the current cross-sectional stuady based on time-series data, the incidence data for confirmed daily cases of COVID-19 from February 26, 2019, to April 25, 2022, were used. A hybrid model based on DWT and ARIMA and a pure ARIMA model were used to predict the trend. All analyzes were performed by MATLAB 2018, stata 2015, and Excel 2013 computer software. ResultsCompared to the ARIMA model, the prediction results of the hybrid model were closer to the actual number of incident cases. The correlation between predicted values by the hybrid model with real data was higher than the correlation between predicted values by the ARIMA model with actual data. ConclusionsDiscreet Wavelet decomposition of the dataset was combined with an ARIMA model and showed better performance in predicting the future trend.

Progressive failure process-considered deformation safety diagnosis method for in-service high arch dam

Conventional methods cannot effectively characterize the progressive failure process, which limits their use in the deformation safety diagnosis of in-service high arch dam. This work proposes a novel method utilizing the frontier theories of mathematics, mechanics, and dam safety monitoring. First, considering the implicit assumption of traditional method, hybrid model (HM) is improved to calibrate the elastic deformation state of high arch dam. Second, the adaptive proportion selection and the neighborhood search strategy are applied to improve artificial bee colony (ABC) algorithm. The undetermined parameters of HM are optimized by the improved ABC. Third, the dangerous degrees of single-point deformation and multi-point deformation are characterized using HM analysis results and information entropy. On this basis, the progressive diagnosis criteria are constructed based on probability principle. Finally, two case studies are conducted to validate the proposed methodology. The analysis results demonstrate that the performance of HM is better than that of the statistical model (SM); the improved ABC promotes the HM performance; and the progressive diagnosis criteria compared with the traditional confidence interval criteria have stricter probabilistic and physical meanings. The 5-level safety control is realized, promoting the emergency response ability of high arch dam in operating condition.

Comparative Analysis of Hybrid Model Performance Using Stacking and Blending Techniques for Student Drop Out Prediction In MOOC

Despite being in high demand as a lifelong learner and academic material supplement, the implementation of Massive Open Online Courses (MOOC) has problems, one of which is the dropout rate (DO) of students, which reaches 93%. As one of the solutions to this problem, machine learning can be utilized as a risk management and early warning system for students who have the potential to drop out. The use of ensemble techniques to build models can improve performance, but previous research has not reviewed the most optimal ensemble technique for this case study. As a form of contribution, this study will compare the performance of models built from stacking and blending techniques. The algorithms used in the base model are KNN, Decision Tree, and Naïve Bayes, while the meta-model uses XGBoost. These algorithms are used to build models with stacking and mixing techniques. The experimental results using stacking are 82.53% accuracy, 84.48% precision, 94.12% recall, and 89.04% F1 score. Meanwhile, the blend obtained 83.39% precision, 85.31% precision, 94.21% recall, and 89.54% F1-Score. These results are supported by model testing using k-fold cross-validation and confusion matrix techniques, which show the same results. That is, blending is 0.86% higher than stacking, so it can be concluded that blending performs better than stacking in the MOOC student dropout prediction case study.

Developing machine learning models with metaheuristic algorithms for droplet size prediction in a microfluidic microchannel

Droplet-based microfluidics holds promise for advanced biomedical applications due to its enhanced accuracy, throughput, rapid reaction time, and minimal reagent consumption. However, achieving droplets with the desired radius often entails costly design iterations. In this paper, we propose a novel approach that combines machine learning algorithms with metaheuristic algorithms to leverage their respective strengths in speed and accuracy. Specifically, we demonstrate the effectiveness of the Rain Optimization Algorithm (ROA) in enhancing the accuracy of five machine learning models. Additionally, we introduce a hybrid model, optimized by ROA, for predicting droplet generation in a cross-junction microfluidic channel, achieving a coefficient of determination (R2) of 0.937, reflecting a 3.6 % increase in accuracy. We employ symbolic regression to validate the hybrid model, yielding an R2 value of 0.987, indicating strong agreement with real data. Furthermore, we utilize the SHAP method in Explainable AI to interpret the hybrid model's performance. Notably, the proposed hybrid model requires only four seconds to solve the prediction problem, significantly reducing computational time compared to the numerical model and experimental method. This approach facilitates the study of each input parameter's impact on droplet size prediction, offering otherwise challenging and costly insights to obtain experimentally.

Enhancing Diabetes Prediction Accuracy through Hybrid Machine Learning Models: A Comparative Study

This study investigates the effectiveness of various machine learning (ML) models in predicting the onset of diabetes, emphasizing the superior performance of hybrid models over single learner models. Employing a dataset comprising 10,000 individuals with features like Glucose level, BMI, Insulin, and more, we meticulously processed and engineered the data to optimize it for ML applications. We developed several models, including Decision Trees, Random Forest, KNN, and XGBoost, and then advanced to hybrid models using ensemble techniques like stacking and soft voting classifiers. Our findings indicate that hybrid models significantly outperform single learner models. These models achieved remarkable accuracy (98.11%), precision (97.31%), and ROC AUC (99.82%), highlighting their potential in clinical settings. The study underscores the value of hybrid ML models in enhancing predictive accuracy and reliability in diabetes diagnostics.

Groundwater level prediction using an improved ELM model integrated with hybrid particle swarm optimisation and grey wolf optimisation

Groundwater, the world's most abundant source of freshwater, is rapidly depleting in many regions due to a variety of factors. Groundwater resources (GWR) play a crucial role in daily life, economic progress, and agricultural crop production, which makes groundwater level (GWL) prediction much more significant for enhanced management. However, one of the major challenges in investigating and estimating groundwater reduction and managing its resources is the absence of complete and reliable data. Therefore, the application of artificial intelligence (AI) models with the necessity for less data and better prediction accuracy is inevitable. The performances of hybrid models including the combination of particle swarm optimisation and grey wolf optimisation with extreme learning machine (ELM-PSOGWO) in estimating the GWL are investigated and evaluated against ELM-GWO, ELM-PSO, ELM, Twin-support vector regression (T-SVR) and SVR models based on the coefficient of determination (R2), mean absolute error (MAE), root mean square error (RMSE), and Nash-Sutclife efficiency coefficient (NSC). Outcomes of this study showed that applied hybridised ELM-PSOGWO (R2 = 0.9969, RMSE = 0.0088, NSC = 0.9961, MAE = 0.0047) performed best, followed by ELM-GWO, ELM-PSO, ELM, TSVR, and SVR in predicting GWL. Based on the obtained results, it is evident that the ELM-PSOGWO model illustrates an excellent agreement with observed GWL data. To this end, this work reveals that hybrid optimisation algorithms can enhance performance of ELM model significantly in GWL prediction. Utilization of advanced AI techniques can well estimate groundwater systems, which saves resources and labor, employed conventionally.

LSTM-CNN Hybrid Model Performance Improvement with BioWordVec for Biomedical Report Big Data Classification

The rise in mortality rates due to leukemia has fueled the swift expansion of publications concerning the disease. The increase in publications has dramatically affected the enhancement of biomedical literature, further complicating the manual extraction of pertinent material on leukemia. Text classification is an approach used to retrieve pertinent and top-notch information from the biomedical literature. This research suggests employing an LSTM-CNN hybrid model to tackle imbalanced data classification in a dataset of PubMed abstracts centred on leukemia. Random Undersampling and Random Oversampling techniques are merged to tackle the data imbalance problem. The classification model’s performance is improved by utilizing a pre trained word embedding created explicitly for the biomedical domain, BioWordVec. Model evaluation indicates that hybrid resampling techniques with domain-specific pre-trained word embeddings can enhance model performance in classification tasks, achieving accuracy, precision, recall, and f1-score of 99.55%, 99%, 100%, and 99%, respectively. The results suggest that this research could be an alternative technique to help obtain information about leukemia.

Comparison of machine learning models for flood forecasting in the Mahanadi River Basin, India

ABSTRACT Developing accurate flood forecasting models is necessary for flood control, water resources and management in the Mahanadi River Basin. In this study, convolutional neural network (CNN) is integrated with random forest (RF) and support vector regression (SVR) for making a hybrid model (CNN–RF and CNN–SVR) where CNN is used as feature extraction technique while RF and SVR are used as forecasting models. These hybrid models are compared with RF, SVR, and artificial neural network (ANN). The influence of training–testing data division on the performance of hybrid models has been tested. Hyperparameter sensitivity analyses are performed for forecasting models to select the best value of hyperparameters and to exclude the nonsensitive hyperparameters. Two hydrological stations (Kantamal and Kesinga) are selected as case studies. Results indicated that CNN–RF model performs better than other models for both stations. In addition, it is found that CNN has improved the accuracy of RF and SVR models for flood forecasting. The results of the training–testing division show that both models’ performance is better at 50–50% data division. Validation results show that both models are not overfitting or underfitting. Results demonstrate that CNN–RF model can be used as a potential model for flood forecasting in river basins.