Stacking Approach Research Articles

Automatic Fake News Identification: A Hybrid CNN-LSTM-Logistic Regression Approach

The news provides important insights into current events and acts as a vital window into the world. The propagation of fake news, poses a serious problem. News that seems to present genuine but is made up is considered fake news. Such fake news can propagate inadvertently or on purpose, foment strife, and erode trust. Identifying fake news has been the focus of various studies to address this issue. To contribute in this direction, we proposed a stacking approach that combines convolutional neural networks (CNN) and long short-term memory (LSTM). We use logistic regression (LR) as a metaclassifier for final classification. We used accuracy, precision, recall, and F1-score as performance evaluation metrics on a real-world dataset. The dataset included in this study reflects a wide range of information and consists of both content from social media platforms and news items from reliable sources. We use McNemar's test to determine the statistical significance of the model's performance. The proposed hybrid approach yields impressive results: 95.19% accuracy, 95.05% precision, 95.54% recall, and 95.29% F1-score. These findings highlight the hybrid model's efficacy in correctly identifying fake news, supporting social peace and the preservation of real news.

A simple unsupervised stacking approach to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity: Using K-means and PCA to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity

A simple unsupervised stacking approach to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity: Using K-means and PCA to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity

A simple unsupervised stacking approach to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity: Using K-means and PCA to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity

A simple unsupervised stacking approach to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity: Using K-means and PCA to Identify in Silico Peptides with Potential Anti-SARS-CoV-2 Activity

Genomic regions associated with spot blotch resistance in elite barley breeding populations

Spot blotch (SB), a prevalent foliar disease of barley, is caused by the hemibiotrophic fungal pathogen Bipolaris sorokiniana. Predominately occurring in humid growing regions worldwide, SB can result in yield losses of up to 30%. Genetic resistance remains the most effective strategy for disease management; however, most Australian barley cultivars exhibit susceptibility despite the previous identification of major resistance loci. This study investigates the genetic architecture underlying spot blotch resistance within an Australian barley breeding program. Resistance was assessed at both the seedling and adult growth stages using a single conidial isolate (SB61) across two consecutive years. A total of 337 barley lines were genotyped with 16,824 polymorphic DArT-seq™ markers. Two mapping approaches were employed: a single-marker genome-wide association study (GWAS) and a haplotype-based local genomic estimated breeding values (Local GEBV) approach. Both methodologies identified two major resistance-associated regions on chromosomes 3H and 7H, effective across growth stages. Additionally, the haplotype-based Local GEBV approach revealed resistance-associated regions on 1H, 3H, and 6H that were not detected by GWAS. Haplotype stacking analysis underscored the critical role of the 7H region for adult-plant resistance when combined with other resistance haplotypes, suggesting significant gene-by-gene interactions and highlighting the complex, quantitative nature of spot blotch resistance. This research confirms the presence of key resistance loci within Australian barley breeding populations, provides novel insight into the genetic architecture of spot blotch resistance, and emphasises the potential to enhance resistance through haplotype stacking and whole-genome prediction approaches.

Enhancing Electricity Consumption Forecasting in Limited Dataset: A Simple Stacked Ensemble Approach Incorporating Simple Linear and Support Vector Regression for Malaysia

Enhancing Electricity Consumption Forecasting in Limited Dataset: A Simple Stacked Ensemble Approach Incorporating Simple Linear and Support Vector Regression for Malaysia

A Novel Ensemble Meta-Model for Enhanced Retinal Blood Vessel Segmentation Using Deep Learning Architectures.

Background: Retinal blood vessel segmentation plays an important role in diagnosing retinal diseases such as diabetic retinopathy, glaucoma, and hypertensive retinopathy. Accurate segmentation of blood vessels in retinal images presents a challenging task due to noise, low contrast, and the complex morphology of blood vessel structures. Methods: In this study, we propose a novel ensemble learning framework combining four deep learning architectures: U-Net, ResNet50, U-Net with a ResNet50 backbone, and U-Net with a transformer block. Each architecture is customized to enhance feature extraction and segmentation performance. The models are trained on the DRIVE and STARE datasets to improve the degree of generalization and evaluated using the performance metric accuracy, F1-Score, sensitivity, specificity, and AUC. Results: The ensemble meta-model integrates predictions from these architectures using a stacking approach, achieving state-of-the-art performance with an accuracy of 0.9778, an AUC of 0.9912, and an F1-Score of 0.8231. These results demonstrate the performance of the proposed technique in identifying thin retinal blood vessels. Conclusions: A comparative analysis using qualitative and quantitative results with individual models highlights the robustness of the ensemble framework, especially under conditions of noise and poor visibility.

PCA-SWF: A Principal Component Analysis-Based Stacked Model Approach for Weather Forecasting

Weather forecasting plays a crucial role across various sectors of society, enabling timely predictions of severe weather events such as hurricanes, tornadoes, storms, floods, and heatwaves. Accurate forecasts provide essential data for issuing public warnings, allowing individuals and authorities to take necessary precautions to safeguard lives and property. In agriculture, farmers rely on weather forecasts to optimize irrigation, planting, and harvesting schedules, ensuring efficient water resource management, maximizing crop yields, and minimizing damage caused by extreme weather. Additionally, weather predictions significantly impact transportation systems by providing insights into potential hazards such as ice, snow, poor visibility, wind speeds, and road conditions. This information helps railways, airlines, and shipping industries adjust schedules and ensure safe and efficient transport operations. In this study, we propose a weather prediction approach using a stacked model, which outperforms single classifiers such as Decision Trees, Support Vector Machines, K-Nearest Neighbors, and Multilayer Perceptrons. Principal Component Analysis (PCA) is employed for feature extraction to reduce dimensionality and improve prediction accuracy. The performance of the proposed model is evaluated using metrics such as accuracy, Matthews Correlation Coefficient (MCC), F1-score, and the Receiver Operating Characteristic (ROC) curve, along with the Area Under the Curve (AUC). The results demonstrate the effectiveness of the stacked model in achieving robust and reliable weather forecasting.

XAI-Enhanced Machine Learning for Obesity Risk Classification: A Stacking Approach with LIME Explanations

XAI-Enhanced Machine Learning for Obesity Risk Classification: A Stacking Approach with LIME Explanations

Implementation Methods for Transmission Zeros in Groove Gap Waveguide Filters at W-Band

In this paper, three groove gap waveguide (GGW) filters are presented, which utilize different techniques to create transmission zeros (TZs) in the W-band. First, a fourth-order filter with cross-coupling of non-adjacent cavities realizes five TZs. The filter is implemented using a stacked cavity approach, which simplifies the realization of cross-couplings with positive and negative signs. The source to load cross-coupling realizes one TZ more than predicted from theory. In the second filter, two off-centered posts with partial height are used to realize two TZs. These posts act as strongly frequency-dependent coupling elements and enable the realization of TZs in an inline filter topology. An interdigital-pin implementation has been used since otherwise, the distance between the posts and the sidewalls is too small to allow manufacturing using the Computerized Numerical Control (CNC) milling approach. The third filter is of fourth order and implements two TZs by using a stub-loaded cavity approach. All filters are designed as a proof of concept in the W-band (75–110 GHz) to verify the manufacturability at high-frequency bands. The manufactured components are realized by high-precision CNC milling and measurement results show good agreement with the simulation for all proposed filters.

Leveraging Stacked Classifiers for Multi-task Executive Function in Schizophrenia Yields Diagnostic and Prognostic Insights.

Cognitive impairment is a central characteristic of schizophrenia. Executive functioning (EF) impairments are often seen in mental disorders, particularly schizophrenia, where they relate to adverse outcomes. As a heterogeneous construct, how specifically each dimension of EF to characterize the diagnostic and prognostic aspects of schizophrenia remains opaque. We used classification models with a stacking approach on systematically measured EFs to discriminate 195 patients with schizophrenia from healthy individuals. Baseline EF measurements were moreover employed to predict symptomatically remitted or non-remitted prognostic subgroups. EF feature importance was determined at the group-level and the ensuing individual importance scores were associated with four symptom dimensions. EF assessments of inhibitory control (interference and response inhibitions), followed by working memory, evidently predicted schizophrenia diagnosis (area under the curve [AUC]=0.87) and remission status (AUC=0.81). The models highlighted the importance of interference inhibition or working memory updating in accurately identifying individuals with schizophrenia or those in remission. These identified patients had high-level negative symptoms at baseline and those who remitted showed milder cognitive symptoms at follow-up, without differences in baseline EF or symptom severity compared to non-remitted patients. Our work indicates that impairments in specific EF dimensions in schizophrenia are differentially linked to individual symptom-load and prognostic outcomes. Thus, assessments and models based on EF may be a promising tool that can aid in the clinical evaluation of this disorder.

HAR: A stacked ensemble learning approach based on smartphone sensors for activity detection

HAR: A stacked ensemble learning approach based on smartphone sensors for activity detection

Non-Invasive Detection of Lung Cancer: Integrating Genomic Data with Ensemble Learning for Improved Early Diagnosis

The early and accurate diagnosis of lung cancer, a significant contributor to global cancer-related mortality, remains a paramount challenge in healthcare. Conventional diagnostic methods often lack the sensitivity required for early-stage detection, prompting the exploration of non-invasive alternatives. Leveraging advancements in genomics and bioinformatics, this study investigates the potential of deoxyribonucleic acid (DNA) and ribonucleic acid (RNA) analysis for early-stage lung cancer diagnosis. Two distinct datasets are utilized: GSE4115, comprising gene expression data from bronchial airway epithelial cells of smokers, and GSE33356, focusing on genomic alterations in Taiwanese female non-smoking lung cancer patients. The research employs comprehensive data pre-processing and feature reduction techniques, including normalization and Kernel Principal Component Analysis (KPCA). Subsequently, an ensemble of diverse learners, including Random Forests, AdaBoost, Bagging, Support Vector Machines (SVMs), and Neural Networks, is trained on the original datasets. A novel ensemble stacking approach is proposed, wherein initial predictions from the base learners are combined through logistic regression - the meta-learner to enhance predictive performance. The study aims to contribute to advancements in lung cancer detection by providing a more precise and non-invasive diagnostic method. By integrating DNA and RNA analysis with ensemble learning techniques, the research endeavours to enhance medical outcomes and potentially save lives through early-stage lung cancer detection.

Prediction of Jakarta's Air Quality Using a Stacking Framework of CLSTM, CatBoost, SVR, and XGBoost

Air quality prediction, particularly in estimating PM10 particle concentration, is a significant challenge in major cities like Jakarta, which experience high levels of air pollution. This study aims to develop an air quality prediction model using an innovative stacking framework that combines several machine learning algorithms, namely ConvLSTM, CatBoost, SVR, and XGBoost. The methodology employed in this research is an experimental approach, where each model is trained and tested individually before being integrated into the stacking framework. The dataset used was sourced from the Kaggle platform, containing historical air quality data in Jakarta. Performance evaluation was conducted by measuring the Root Mean Squared Error (RMSE) for each model. The results of the study showed that the ConvLSTM model produced an RMSE of 13.5168, CatBoost with an RMSE of 13.4113, and SVR with an RMSE of 14.2725. To improve prediction accuracy, the researchers employed a stacking approach of the four models (ConvLSTM, CatBoost, SVR, and XGBoost), which yielded a much lower RMSE of 0.8093. Thus, this stacking framework has proven to significantly enhance air quality prediction performance, particularly in predicting PM10 concentrations in Jakarta.

Classification of Mental Health Care Using the ELM, MLP, and CatBoost Stacking Framework

Mental health significantly impacts overall well-being, yet the increasing prevalence of mental health issues presents challenges in their effective classification and treatment. Traditional methods often fail to accurately handle complex, non-linear data, compromising the timeliness and appropriateness of interventions. This study introduces an innovative mental health classification framework, ELM-MLP-CatBoost Stacking, to address these deficiencies. The primary objective is to enhance classification accuracy by integrating three advanced computational techniques: the speed of the Extreme Learning Machine (ELM), the flexibility of the Multi-Layer Perceptron (MLP) for modeling non-linear data, and the predictive refinement of CatBoost as a meta-model. Our methodology involves a stacking approach where ELM and MLP models serve as base learners with CatBoost integrating their outputs to optimize final predictions. Experimental results demonstrate that the ELM-MLP-CatBoost Stacking framework substantially outperforms traditional models, achieving a notable accuracy of 92.76%, an improvement over the MLP’s 92.64% and the ELM’s 69.59%. This framework enhances the reliability and efficiency of mental health condition classifications and paves the way for further research into advanced diagnostic tools. The novelty of this research lies in the synergistic combination of these models, setting a new standard for accuracy and reliability in mental health diagnostics and establishing a robust foundation for future advancements in the field.

Web-Based Task Organizer for Enhanced Productivity

The Web-Based Task Organizer for Enhanced Productivity (WBTO) mainly aims to improve task management and overall increase in productivity. In our system we offer features such as task prioritization, categorization, due dates and recurrence options which is integrated into a user friendly interface. We use a full stack approach using java for the back end logic, JavaScript for a dynamic user interface and a SQL database for preserving data storage. It provides key features like task creation, categorization, prioritization, scheduling and progress tracking and interactive functionalities and ensuring data integrity and privacy. As this not only improves individual productivity but also further provides group work collaboration allowing for much faster task tracking, reminders and analytics which improves overall user experience.

A Stacked Neural Network Model for Damage Localization.

Traditional vibration-based damage detection methods often involve human intervention in decision-making, therefore being time-consuming and error-prone. In this study, we propose using Artificial Neural Networks (ANNs) to detect patterns in the structural response and create accurate predictions. The features extracted from the response signal are the Relative Frequency Shifts (RFSs) of the first eight weak-axis bending vibration modes, and the predictions refer to the damage location. To increase the accuracy of the predictions, we propose a novel stacked neural network approach, capable of detecting damage locations with high accuracy. The dataset used for training involves, as input data, the RFSs calculated with an original method for numerous damage locations and severities. The following models were used as building blocks for our stacked approach: Multilayer Perceptron, Recurrent Neural Network, Long Short-term Memory, and Gated Recurrent Units. The entire beam was thus split into segments and each network was trained in this stacked model on one beam segment. All results obtained with the models are also compared to a standard neural network trained on the entire beam. The results obtained show that the model that performs the best contains 14 stacked two-layer feedforward networks.

An ensemble machine learning framework with explainable artificial intelligence for predicting haemoglobin anaemia considering haematological markers

Anaemia is a disorder marked by low blood levels of haemoglobin (HGB), affecting people of all ages and ethnicities and is a major global public health concern. Anaemia must be diagnosed as soon as possible to enable prompt treatment and intervention, which can reduce complications and enhance patient outcomes. With the ability to improve diagnostic precision and expedite patient care procedures, machine learning (ML) has become a potent instrument in the healthcare industry. Hence, we examine the use of ML approaches to predict haemoglobin-like anaemia in this research article. Based on a heterogeneous dataset of blood markars, we investigate the performance of many machine learning techniques such as Logistic Regression, CatBoost, XgBoost Decision Trees, KNN and others. The algorithms are further ensembled using a customized stacking approach. The ML models' judgments are interpreted using explainable artificial intelligence (XAI) methods. The xgboost and the stacking classifier obtained an accuracy, precision and recall of 99% each. Our research shows how ML models can help with the early diagnosis and treatment of anaemia, which will ultimately lead to better patient outcomes and healthcare results. Overall, the research shows how ML emphasizes the value of interdisciplinary cooperation in solving challenging medical problems.

Balancing data imbalance in biomedical datasets using a stacked augmentation approach with STDA, DAGAN, and pufferfish optimization to reveal AI's transformative impact

Balancing data imbalance in biomedical datasets using a stacked augmentation approach with STDA, DAGAN, and pufferfish optimization to reveal AI's transformative impact

Vertical stacking approach for tailoring electronic and optical properties of ultrathin two-dimensional vdW heterostructures of P-MAB (M=Ti, Zr, Hf; A=S, B=Se)

The vertical stacking in the form of heterostructures, is an effective way to enhance the performance of corresponding isolated monolayers. Strong interactions between Janus monolayers (MAB) and Phosphorus (P) monolayers result in significant improvements in both physical and chemical properties compared to single layer configurations. Using first-principle (hybrid) calculations, we explore electronic structure and optical properties of P–MAB (M = Ti, Zr, Hf; A = S, B=Se) van der Waals heterostructures, for potential applications in electronic and optoelectronic devices. Ab initio molecular dynamic (AIMD) simulations confirm the thermal stability, while phonon spectra confirm the dynamical stabilities of these systems. Our analysis reveals that all configurations maintain semiconducting nature with an indirect bandgap. The type-II band alignment in the P–MAB heterostructures is demonstrated through calculations of the partial density of states (PDOS). Bader charge analysis and planar-averaged charge density differences indicate an electron transfer from the P monolayers to the MAB monolayer. Furthermore, our investigation of the dielectric function ϵ2(ω) shows that the lowest energy shift are primarily influenced by excitonic effects.

Van der Waals Heterojunction Based Self-Powered Biomimetic Dual-Mode Sensor for Precise Object Identification.

The design and fabrication of materials that can concurrently respond to light and gas within the dual-modal recognition domain present a significant challenge due to contradictory structural requirements. This innovative strategy introduces a type-I heterojunction, combining the properties of Sb2Te3 and WSe2 nanosheets, to overcome these obstacles. The heterojunction is prepared through a precise stacking approach to create a single-side barrier on the valence band and a near-zero offset on the conduction band. The resulting Sb2Te3/WSe2 heterojunction demonstrates unparalleled performance, showcasing the best integrated photoelectric and gas sensing performance in a single device to date. Based on the above features, the heterojunction successfully integrates visual and olfactory sensing performance, achieving the first biomimetic visual-olfactory dual-mode recognition in a single device. This simulation increased the accuracy of distinguishing electric and fuel-powered cars from ≈50% to ≈96%. This work introduces a novel approach to creating efficient, self-powered sensing materials, paving the way for next-generation biomimetic dual-model devices with broad applications in environmental protection, medical care, and other fields.