Stacking Technique Research Articles

Optimization of Cocoa Pods Maturity Classification Using Stacking and Voting with Ensemble Learning Methods in RGB and LAB Spaces

Determining the maturity of cocoa pods early is not just about guaranteeing harvest quality and optimizing yield. It is also about efficient resource management. Rapid identification of the stage of maturity helps avoid losses linked to a premature or late harvest, improving productivity. Early determination of cocoa pod maturity ensures both the quality and quantity of the harvest, as immature or overripe pods cannot produce premium cocoa beans. Our innovative research harnesses artificial intelligence and computer vision technologies to revolutionize the cocoa industry, offering precise and advanced tools for accurately assessing cocoa pod maturity. Providing an objective and rapid assessment enables farmers to make informed decisions about the optimal time to harvest, helping to maximize the yield of their plantations. Furthermore, by automating this process, these technologies reduce the margins for human error and improve the management of agricultural resources. With this in mind, our study proposes to exploit a computer vision method based on the GLCM (gray level co-occurrence matrix) algorithm to extract the characteristics of images in the RGB (red, green, blue) and LAB (luminance, axis between red and green, axis between yellow and blue) color spaces. This approach allows for in-depth image analysis, which is essential for capturing the nuances of cocoa pod maturity. Next, we apply classification algorithms to identify the best performers. These algorithms are then combined via stacking and voting techniques, allowing our model to be optimized by taking advantage of the strengths of each method, thus guaranteeing more robust and precise results. The results demonstrated that the combination of algorithms produced superior performance, especially in the LAB color space, where voting scored 98.49% and stacking 98.71%. In comparison, in the RGB color space, voting scored 96.59% and stacking 97.06%. These results surpass those generally reported in the literature, showing the increased effectiveness of combined approaches in improving the accuracy of classification models. This highlights the importance of exploring ensemble techniques to maximize performance in complex contexts such as cocoa pod maturity classification.

Enhancing Multiple Linear Regression with Stacking Ensemble for Dissolved Oxygen Estimation

Maintaining optimal dissolved oxygen levels is essential for aquatic ecosystems, yet industrial and domestic waste has led to a global decline in dissolved oxygen. Traditional measurement methods, such as oxygen meters and Winkler titration, are often costly or time-consuming. This study aims to improve the Root Mean Square Error, Mean Absolute Error, and R2 values for estimating dissolved oxygen levels. The research method uses Multiple Linear Regression with various training and testing data splits, both before and after applying polynomial features. The model is further optimized using a stacking technique, with Random Forest Regressor and Gradient Booster Regressor as base models.The results show that the best model was achieved using the stacking ensemble technique with a 90:10 data split and polynomial features, yielding a Root Mean Square Error of 1.206, Mean Absolute Error of 0.990, and R2 of 0.670. This model has also met the assumptions of linear regression, such as residual normality, homoscedasticity, and no autocorrelation of residuals. This study concluded that the ensemble stacking technique and the addition of polynomial features could improve the model in estimating dissolved oxygen values and also contribute by providing an accessible user interface using the Gradio Framework, allowing users to estimate dissolved oxygen levels effectively.

Out-of-plane polarization induces a picosecond photoresponse in rhombohedral stacked bilayer WSe2.

Constructing van der Waals materials with spontaneous out-of-plane polarization through interlayer engineering expands the family of two-dimensional ferroelectrics and provides an excellent platform for enhancing the photoelectric conversion efficiency. Here, we reveal the effect of spontaneous polarization on ultrafast carrier dynamics in rhombohedral stacked bilayer WSe2. Using precise stacking techniques, a 3R WSe2-based vertical heterojunction was successfully constructed and confirmed by polarization-resolved second harmonic generation measurements. Through output characteristics and the scanning photocurrent map under zero bias, we reveal a non-zero short-circuit current in the graphene/3R WSe2/graphene heterojunction region, demonstrating the bulk photovoltaic effect. Furthermore, the out-of-plane polarization enables the 3R WSe2 heterojunction region to achieve an ultrafast intrinsic photoresponse time of approximately 3 ps. The ultrafast response time remains consistent across varying detection powers, demonstrating environmental stability and highlighting the potential in optoelectronic applications. Our study presents an effective strategy for enhancing the response time of photodetectors.

Ensemble Stacking of Machine Learning Approach for Predicting Corrosion Inhibitor Performance of Pyridazine Compounds

Corrosion is a major challenge affecting various industrial sectors, leading to increased operational costs and decreased equipment efficiency. The use of organic corrosion inhibitors is one of the promising solutions. This study applies an ensemble algorithm with a stacking method to estimate pyridazine-derived compounds corrosion inhibition efficiency. This study utilized various molecular characteristics of pyridazine compounds as inputs to predict inhibition efficiency values. After evaluating several boosting models, the stacking technique was chosen as it showed the best results. Stacking Model 6, which combines XGB, LGBM, and CatBoost as the base model with Random Forest as the meta-model, produced the most accurate prediction with an RMSE of 0.055. These findings indicate that machine learning approaches can effectively and efficiently predict corrosion inhibitor performance. This method offers a faster and more economical alternative to conventional experimental methods.

Classification of Heart Disease Using a Stacking Framework of BiGRU, BiLSTM, and XGBoost

This study aims to develop a heart disease classification model using an ensemble approach by leveraging a Stacking framework that combines BiGRU, BiLSTM, and XGBoost models. In this research, the BiGRU and BiLSTM models are utilized as base models to extract temporal and spatial features from sequential data, while XGBoost is employed as a metamodel to perform the final classification based on the features generated by the two base models. The test results show that the BiGRU model achieves an accuracy of 0.77, while the BiLSTM model achieves an accuracy of 0.85. By applying the Stacking technique using XGBoost as the meta-model, the classification accuracy significantly increases to 0.92. These findings indicate that the Stacking framework can effectively enhance heart disease classification performance, making it a potentially powerful tool for medical applications in heart disease diagnosis.

High-Fidelity OC-Seislet Stacking Method for Low-SNR Seismic Data

Seismic stacking is a core technique in seismic data processing, aimed at enhancing the signal-to-noise ratio (SNR) of data by utilizing seismic data acquisition with multifold geometry. Traditional stacking methods always have certain limitations, such as the reliance on the accuracy of velocity analysis for dip moveout (DMO) in common midpoint (CMP) stacking. The seislet transform, a compression technique tailored to nonstationary seismic data, can compress and stack along the prediction direction of seismic data, which provides a new technical idea for high-fidelity seismic imaging based on the effectiveness of the compression. This paper introduces a high-order OC-seislet stacking method for low-SNR seismic data, capable of achieving the high-fidelity stacking of reflection and diffraction waves simultaneously. With the multi-scale analysis advantages of the seislet transform, this method addresses the dependency of DMO stacking on velocity analysis accuracy. In the frequency–wavenumber–scale domain, the correction compensation of the high-order CDF 9/7 basis function is used to obtain the compression coefficients of the high-order OC-seislet transform. This approach simultaneously stacks frequency–wavenumber information of reflection and diffraction waves with high fidelity while implementing DMO processing. After normalizing the weighting coefficients and applying soft thresholding for denoising, the final result is transformed back to the original time–space domain, yielding high-fidelity stacking sections. The results of applying this method to both synthetic and field data show that, compared with conventional DMO stacking methods, the high-order OC-seislet stacking technique reasonably represents dipping layers and fault amplitudes, and it can achieve a balance of a high SNR and high fidelity in stacked profiles.

Tunnel-induced land subsidence assessment in a densely populated residential area using Sentinel-1 PS-InSAR

Land subsidence detection in residential areas based on advanced remotely sensed data is an effective technique to mitigate catastrophic disasters in time. Synthetic aperture radar (SAR) in spaceborne platforms is a highly sophisticated technique to estimate ground surface subsidence with millimeter accuracy for wide-area observation. SAR is beneficial for large-scale social infrastructure project planning and monitoring in densely populated residential areas. In this study, we applied a permanent scatterer (PS) interferometric SAR (PS-InSAR)-based technique for monitoring land subsidence in a tunnel construction site in the Greater Tokyo Area to investigate tunnel-induced ground surface subsidence. Sentinel-1 repeat-pass SAR data in ascending and descending orbital directions were used for continuous observation. Consecutively estimated interferograms based on repeat-pass SAR data were generated to monitor and detect tunnel-induced land subsidence as well as confirm long-term stability along the tunnel excavation path. A bidirectional displacement of the ground area parallel to the earth’s surface in the east–west direction and perpendicular to the earth’s surface in the up–down direction was estimated at a high-phase consistency location using a PS interferometric stacking technique. The resultant data were then analyzed to detect anomalies in the ground area in the spatiotemporal domains by performing displacement time-series analysis. An area near Higashi-Tsutsujigaoka’s local administrative region in the metropolitan center of Tokyo was observed to have 15-mm subsidence and 30-mm displacement in the east–west direction during August–October 2020. The conventional level survey results in the same ground area showed 15–20-mm subsidence above the tunnel excavation path. The findings of this study offer insights into understanding land subsidence in residential environments.

Monte‐Carlo Study on Temperature Effect on Scanning Electron Microscopy Contrast for One Spherical Nanoparticle in Multiframe Stacking Mode

Modern scanning electron imaging (SEM) has become an essential tool in various fields. In order to obtain high‐quality images with minimal vibration and drift, the multi‐frame stacking technique is commonly used. However, one aspect that has been overlooked in this process is the electron beam‐induced temperature effect. This study aims to explore the impact of the temperature effect on SEM contrast for a single spherical nanoparticle, specifically a gold nanosphere on a carbon substrate. Through Monte‐Carlo simulation, we investigated how the number of frames in multi‐frame stacking mode affects two types of electron signals: secondary electrons (SE) and backscattering electrons (BSE). Our findings revealed that the SEM contrast is indeed sensitive to the number of frames used in the stacking process. It is observed that the SE contrast decreases while the BSE contrast increases compared to cases without considering temperature effects. It is also found that when using the large primary electron beam energy, the impact of temperature effect was less significant. The electron‐solid interaction theory was utilized to systematically explain the underlying mechanism. This study not only sheds light on how temperature effects can influence SEM imaging but also provides valuable insights for optimizing image acquisition techniques.

Identifying Potential Landslides in Low-Coherence Areas Using SBAS-InSAR: A Case Study of Ninghai County, China

The southeastern coastal regions of China are characterized by typical hilly terrain with abundant rainfall throughout the year, leading to frequent geological hazards. To investigate the measurement accuracy of surface deformation and the effectiveness of error correction methods using the small baselines subset–interferometry synthetic aperture radar (SBAS-InSAR) method in identifying potential geological hazards in such areas, this study processes and analyzes 129 SAR images covering Ninghai County, China. By processing coherence coefficients using the Stacking technique, errors introduced by low-coherence images during phase unwrapping are mitigated. Subsequently, interferograms with high coherence are selected for time-series deformation analysis based on the statistical parameters of coherence coefficients. The results indicate that, after mitigating errors from low-coherence images, applying the SBAS-InSAR method to only high-coherence SAR datasets provides reliable surface deformation results. Additionally, when combined with field geological survey data, this method successfully identified landslide boundaries and potential landslides not accurately detected in previous geological surveys. This study demonstrates that using the SBAS-InSAR method and selecting high-coherence SAR images based on interferogram coherence statistical parameters significantly improves measurement accuracy and effectively identifies potential geological hazards.

The hot circumgalactic medium in the eROSITA All-Sky Survey

Context. The circumgalactic medium (CGM) provides the material needed for galaxy formation and influences galaxy evolution. The hot (T > 106K) CGM is poorly detected around galaxies with stellar masses (M*) lower than 3 × 1011 M⊙ due to the low surface brightness. Aims. We aim to detect the X-ray emission from the hot CGM around Milky Way-mass (MW-mass, log(M*/M⊙) = 10.5 − 11.0) and M31-mass (log(M*/M⊙) = 11.0 − 11.25) galaxies, in addition to measuring the X-ray surface brightness profile of the hot CGM. Methods. We applied a stacking technique to gain enough statistics to detect the hot CGM. We used the X-ray data from the first four SRG/eROSITA All-Sky Surveys (eRASS:4). We discussed how the satellite galaxies could bias the stacking and the method we used to carefully build the central galaxy samples. Based on the SDSS spectroscopic survey and halo-based group finder algorithm, we selected central galaxies with spectroscopic redshifts of zspec < 0.2 and stellar masses of 10.0 < log(M*/M⊙) < 11.5 (85 222 galaxies) – or halo masses of 11.5 < log(M200m/M⊙) < 14.0 (125,512 galaxies). By stacking the X-ray emission around galaxies, we obtained the mean X-ray surface brightness profiles. We masked the detected X-ray point sources and carefully modeled the X-ray emission from the unresolved active galactic nuclei (AGN) and X-ray binaries (XRB) to obtain the X-ray emission from the hot CGM. Results. We measured the X-ray surface brightness profiles for central galaxies of log(M*/M⊙) > 10.0 or log(M200m/M⊙) > 11.5. We detected the X-ray emission around MW-mass and more massive central galaxies extending up to the virial radius (Rvir). The signal-to-noise ratio (S/N) of the extended emission around MW-mass (M31-mass) galaxy is about 3.1σ (4.7σ) within Rvir. We used a β model to describe the X-ray surface brightness profile of the hot CGM (SX, CGM). We obtained a central surface brightness of log(SX,0[erg s−1 kpc−2]) = 36.7−0.4+1.4 (37.1−0.4+1.5) and β = 0.43−0.06+0.10 (0.37−0.02+0.04) for MW-mass (M31-mass) galaxies. For galaxies with log(M200m/M⊙) > 12.5, the extended X-ray emission is detected with S/N > 2.8σ and the SX, CGM can be described by a β model with β ≈ 0.4 and log(SX,0[erg s−1 kpc−2]) > 37.2. We estimated the baryon budget of the hot CGM and obtained a value that is lower than the prediction of ΛCDM cosmology, indicating significant gas depletion in these halos. We extrapolated the hot CGM profile measured within Rvir to larger radii and found that within ≈3Rvir, the baryon budget is close to the ΛCDM cosmology prediction. Conclusions. We measured the extended X-ray emission from representative populations of central galaxies around and above MW-mass out to Rvir. Our results set a firm footing for the presence of the hot CGM around such galaxies. These measurements constitute a new benchmark for galaxy evolution models and possible implementations of feedback processes therein.

Machine Learning for Enhanced Churn Prediction in Banking: Leveraging Oversampling and Stacking Techniques

Every sector of business is getting more competitive as time passes. More and more companies are offering services to people. Banking sector is no different. With the plethora option customer has in terms banking, holding on to customer may prove difficult for banks. This research will help banks to predict which customers are likely going to churn and allow them to take precaution to stop customers from leaving. In this study we have used classifiers such as: K-Neighbors, Random Forest, XGboost, Adaboost classifiers and Ensemble Model (Stacking Technique) that uses all of these models together. The experimentation was conducted on a dataset from Kaggle. The dataset used in this research was heavily imbalanced. So, different oversampling methods like Random Oversampling and SMOTE-ENN have been used. In data preprocessing, label encoding was done and for validation K-folding technique (k-5) have been used. The highest accuracy has been achieved by using Random Oversampling with Stacking Model which is 97.31% (std: 0.0033, k=5).

The SRG/eROSITA all-sky survey. X-ray emission from the warm-hot phase gas in long cosmic filaments

The properties of the warm-hot intergalactic medium (WHIM) in cosmic filaments are among the least quantified units in modern astrophysics. The Spectrum Roentgen Gamma/eROSITA All Sky Survey ((SRG/eRASS) provides a unique opportunity to study the X-ray emission of the WHIM. We applied both imaging and spectroscopic stacking techniques to the data of the first four eRASS scans to inspect the X-ray emissions from 7817 cosmic filaments identified from Sloan Digital Sky Survey (SDSS) optical galaxy samples. We obtained a $9 significant detection of the total X-ray signal from filaments in the 0.3--1.2 keV band. Here, we introduce a novel method to estimate the contamination fraction from unmasked X-ray halos, active galactic nuclei, and X-ray binaries associated with filament galaxies. We found an approximately 40<!PCT!> contamination fraction for these unmasked sources, suggesting that the remaining 60<!PCT!> of the signal could be coming from the WHIM and a $5.4 detection significance of the WHIM. Moreover, we modeled the temperature and baryon density contrast of the detected WHIM by fitting the stacked spectrum and surface brightness profile. The best-fit temperature $ K obtained by using a single temperature model, is marginally higher than in the simulation results. This could be due to the fitting of a single temperature model on a multi-temperature spectrum. Assuming a 0.2 solar abundance, the best-fit baryon density contrast $ b is in general agreement with the X-ray emitting phases in the IllustrisTNG simulation. This result suggests that the broadband X-ray emission traces the high end of the temperature and density values that characterize the entire WHIM population.

Materials With the Help of Mechanical Properties for Electrical Vehicle Chassis using Machine Learning Methods

Material science is a fast-growing research field where artificial intelligence is applied in a variety of applications to provide accurate solutions to the problem. Due to its generalizability, noise tolerance, and fast computation, machine learning algorithms have emerged in recent years as a potent tool for creating correlations between data, and are finding use in materials science. In this research work electric vehicle chassis material selection is done based on the mechanical properties of the material and this is done using machine learning techniques. Machine learning techniques, like logistic regression, K-Nearest Neighbor, Decision Tree, Random Forest, Naïve Base, XGBoost and AdaBoost techniques are used for the same. The stacking technique is also used which combines a variety of ML algorithms for enhanced performance and is observed that the stacking technique gives better accuracy compared to other classifiers. Binary class, as well as multiclass problems, are taken that will give solutions to the electric vehicle chassis selection material. Accuracy scores of different algorithms are compared and found that stacking works reasonably better compared to others.

Bias correction and ensemble techniques in statistical downscaling model for rainfall prediction using Tweedie-LASSO in West Java, Indonesia

Rainfall is a climate element with high variations in space and time scales, so it is not easy to predict. One way to predict rainfall is statistical downscaling (SD). SD can predict local rainfall based on Global Circulation Model (GCM) data. The Decadal Climate Prediction Project (DCPP), one of the GCMs, originates from adjacent grids and experiences multicollinearity problems. Rainfall as a response variable is Tweedie Compound Poisson Gamma (TCPG) distribution data because it has a discrete component (rainfall events) and a continuous component (rainfall intensity), so SD modelling will be carried out using Tweedie-LASSO. This research aims to compare the performance of bias correction and ensemble methods in SD in predicting rainfall in West Java, Indonesia. Bias correction uses Empirical Quantile Mapping (EQM) with CHIRPS data, and the ensemble method uses a stacking technique with Random Forest (Stacking-RF) due to the varied characteristics of DCPP model sources. Evaluation results using Root Mean Square Error Prediction (RMSEP) and correlation coefficient show that bias correction improves single-model performance but not ensemble models. Besides that, ensemble models outperform single models both before and after bias correction. The combination of bias correction and ensemble modelling can be recommended when conducting SD to enhance the prediction capability of rainfall at stations and other areas.

Identification of millet origin using terahertz spectroscopy combined with ensemble learning

It’s crucial for both producers and consumers to accurately trace the origin of millet, given the significant differences in price and taste that exist between millets from various origins. The traditional method of identifying the origin of millet is time-consuming, laborious, complex, and destructive. In this study, a new method for fast and non-destructive differentiation of millet origins is developed by combining terahertz time domain spectroscopy with ensemble learning. Firstly, three machine learning algorithms, namely support vector machine (SVM), random forest (RF), and kernel extreme learning machine (KELM), were used to build different discriminative models, and then the impact of six different preprocessing methods on the models’ classification performance was compared. It was observed that models employing Savitzky-Golay preprocessing exhibited pronounced superiority in accurately determining the millet’s geographical origins. Building upon these findings, the research introduces an innovative ensemble learning strategy, leveraging both topsis and stacking techniques, to harness the collective strengths of the three algorithms. The outcomes of this approach reveal its remarkable capacity to distinguish millets originating from five distinct locations without the necessity for any parameter fine-tuning. The accuracy, F1 score, and Kappa on the prediction set are all 100 %, which significantly outperforms the single model, traditional voting method, and stacking method. The culmination of this study suggests that the integration of terahertz time-domain spectroscopy and TOPSIS-Stacking ensemble learning emerges as a promising method for the swift and non-intrusive discrimination of millet geographical origins with remarkable precision.

Advancements in Multiple‐Step On‐Line Preconcentration Techniques for Enhanced Sensitivity in Capillary Electrophoresis

ABSTRACTMultiple‐step on‐line preconcentration, a combination of at least two stacking techniques has been developed to increase the sensitivity in capillary electrophoresis (CE) for analytes in various samples. It is usually conducted sequentially, or in some cases, synergistically, where different stacking modes occur simultaneously. Multiple‐step techniques allow simultaneous preconcentration and separation of various kinds of analytes in different complex samples in a single CE run. This review aims to provide recent advances in multiple‐step on‐line preconcentration techniques in CE. We critically review technical papers published for the last 7 years up until July 2024, subsequently organized according to the combination of the main stacking techniques, that is, field amplification, large volume sample stacking, transient isotachophoresis, micelle to solvent or micelle to cyclodextrin stacking, and others. The procedures, fundamental mechanism, analytical figures of merits achieved, and their feasibility for complicated sample matrices are reviewed.

An advanced double-phase stacking ensemble technique with active learning classifier: Toward reliable disruption prediction in Aditya tokamak.

Disruptions in tokamak nuclear reactors, where plasma confinement is suddenly lost, pose a serious threat to the reactor and its components. Classifying discharges as disruptive or non-disruptive is crucial for effective plasma operation and advanced prediction. Traditional disruption identification systems often struggle with noise, variability, and limited adaptability. To address these challenges, we propose an enhanced stacking generalization model called the "Double-Phase Stacking Technique" integrated with Pool-based Active Learning (DPST-PAL) for designing a robust classifier with minimal labor cost. This innovative approach improves classification accuracy and reliability using advanced data analysis techniques. We trained the DPST-PAL model on 162 diagnostic shots from the Aditya dataset, achieving a high accuracy of 98% and an F1-score of 0.99, surpassing conventional methods. Subsequently, the deep 1D convolutional predictor model is implemented and trained using the classified shots obtained from the DPST-PAL model to validate the reliability of the dataset, which is tested on 47 distinct shots. This model accurately predicts the disruptions 7-13ms in advance with 93.6% accuracy and exhibited no premature alarms or misclassifications for our experimental shots.

Enhancing Accuracy by Using Boosting and Stacking Techniques on the Random Forest Algorithm on Data from Social Media X

Enhancing Accuracy by Using Boosting and Stacking Techniques on the Random Forest Algorithm on Data from Social Media X

Enhancement-mode ZnGa2O4-based Phototransistor with high UV–visible rejection ratio grown by metalorganic chemical vapor deposition

In this study, a high-crystallinity zinc gallium oxide (ZnGa2O4) epilayer of about 80 nm thickness was successfully grown using metalorganic chemical vapor deposition. X-ray diffraction, atomic force microscopy, UV–visible spectroscopy measurements, and X-ray photoelectron spectroscopy were used to investigate the quality of the epilayer. Transmission line measurement analysis revealed the electrical properties of the ZnGa2O4 epilayer: a sheet resistance of 276 MΩ/□, a transfer length of 0.28 μm, and a specific contact resistivity of 21.63 Ω/mm2. A phototransistor made of ZnGa2O4 was designed with a channel width of 250 μm and a channel length of 20 μm. The source-drain electrodes were fabricated using a Ti/Al/Ni stacking technique. The gate dielectric comprised 30 nm thick Al2O3, and the gate contact was made of Ni. The device possessed a higher threshold voltage value of about 4.8 V and an impressive On/Off current ratio of 106. Under 240 nm UV light exposure, the phototransistor exhibited a transition from enhancement mode to depletion mode, leading to a remarkable shift in the threshold voltage from 5 V to approximately −16 V. Moreover, the responsivity and rejection ratio at 240 nm was 128.5 A/W and 105 at a gate voltage of −10 V. The study shows that an increasing thickness of the ZnGa2O4 epilayer from 80 nm to 135 nm results in a depletion mode transistor. The higher concentration of about twenty percent oxygen vacancies resulted in a longer fall time for the phototransistor. These results underscore the significant potential of ZnGa2O4 in facilitating deep UV-based detection applications.

Extreme flash flood susceptibility mapping using a novel PCA-based model stacking approach

This study introduces an efficient methodology for model stacking, incorporating six diverse machine learning and statistical models alongside principal component analysis (PCA). The approach is applied for the flash flood susceptibility mapping within the Karkheh Basin in Iran. The selected models include random forest (RF), boosted regression trees (BRT), support vector machine (SVM), artificial neural networks (ANN), generalized additive model (GAM), and the least absolute shrinkage and selection operator (Lasso), with RF also serving as the meta-model for the stacking. The results revealed significant correlations among the predictions of the individual models, which could potentially impact the meta-model’s efficacy. To address this, PCA was applied to the model predictions to generate de-correlated components as inputs for the meta-model, thereby enhancing prediction accuracy and robustness. Evaluation based on the area under the receiver operating characteristic (AUROC) curve demonstrated that the GAM outperformed all other individual models with the highest accuracy score of 0.924. In contrast, the RF and ANN models had the lowest accuracy, both registering at 0.872. However, the performance disparity across models was minimal. Notably, the PCA-based stacking approach (0.936) surpassed both traditional model stacking (0.912) and the performances of all individual models, advocating for its use in enhancing predictive accuracy. These findings endorse the PCA-stacking method over conventional stacking techniques. Nonetheless, further research across varied applications is warranted to generalize its efficacy.