DADaniel Cristóbal Andrade-Girón

Daniel Cristóbal Andrade-Girón

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Daniel Cristóbal Andrade-Girón has a total of 17 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Leakage-Aware Evaluation of Ensemble Classifiers for Symptom-Based Classification of Diabetes Status: Performance, Calibration, and Explainability

    by Daniel Andrade-Girón, William Marin-Rodriguez, Americo Peña, Enrique Diaz-Vega, Abrahán Neri-Ayala, Viviana Vellon-Flores, Timoteo Solano-Armas, Algemiro Muñoz-Vilela

    Background/Objectives: Diabetes impose a substantial public health burden. Machine learning models trained on symptom data may help identify individuals whose recorded symptom profiles warrant confirmatory clinical assessment. This study compared ten en-semble models for symptom-based classification of a dataset-provided diabetes-status la-bel. Methods: The UCI Early-Stage Diabetes Risk Prediction Dataset was used, comprising 520 observations, 251 unique predictor profiles, 16 predictors, and a binary positive/negative diabetes-status label. All observations were retained, while identical predictor profiles were kept within the same partition during data splitting, nested cross-validation, hy-perparameter optimization, and calibration to reduce information leakage. The weighted F1-score was the optimization objective and primary model-selection metric throughout the analysis. Results: Bagging Extra Tree achieved the highest observed mean weighted F1-score (94.07% ± 7.99%) and MCC (0.876 ± 0.167) and was selected as the final model under the prespecified criterion. An exploratory Friedman analysis indicated variation in model rankings across folds (p = 0.001859), although no Nemenyi-adjusted pairwise comparison reached the nominal significance level. On the profile-disjoint internal holdout set, the calibrated model achieved a weighted F1-score of 0.9224, an MCC of 0.8207, a ROC-AUC of 0.9844, and a Brier score of 0.0528. Random Forest was selected in the unique-profile sensitivity analysis but yielded lower holdout performance, indicating sensitivity to the analytical unit. Conclusions: These findings provide internal evidence that tree-based ensembles can clas-sify the dataset-provided diabetes-status label under profile-disjoint evaluation. They do not establish prospective risk prediction or clinical diagnostic utility; independent exter-nal validation is required before clinical implementation.

  • Open Access

    Predicting Student Stress Using Machine Learning Ensemble Models: A Multi-Criteria Comparison with Explainable Artificial Intelligence Analysis

    by Daniel Cristóbal Andrade-Girón, William Joel Marin-Rodriguez, Marcelo Gumercindo Zuñiga-Rojas, Abrahan Cesar Neri-Ayala, Edgar Tito Susanibar-Ramírez, Miguel Angel Aguilar-Luna-Victoria

    Student stress is a significant mental health issue in educational settings; therefore, developing reliable, calibrated, and interpretable predictive models can support the classification of observed stress levels. This study analyzed the public Student Stress Factors dataset, comprising 1100 records, 20 predictors, and one target variable, using a supervised machine learning pipeline designed to reduce information leakage. The pipeline included stratified data partitioning, encapsulated preprocessing, nested cross-validation restricted to the training data, and independent holdout evaluation. Nine ensemble and boosting algorithms for tabular data were compared: AdaBoost, Gradient Boosting, Random Forest, Extra Trees, Bagging, Voting, Stacking, XGBoost, and LightGBM. Model performance was assessed using key discrimination and calibration metrics, together with the nonparametric Friedman test for statistical comparison. Gradient Boosting achieved the best average performance in nested cross-validation, with an accuracy of 89.55 ± 3.16%, F1-weighted of 89.54 ± 3.17%, MCC of 0.845 ± 0.047, and ROC-AUC weighted of 98.59 ± 0.92%. XGBoost and LightGBM showed comparable performance. In the independent holdout set, the final calibrated model maintained robust predictive performance, achieving an accuracy of 0.8818, F1-weighted of 0.8818, MCC of 0.8237, and ROC-AUC weighted of 0.9861. Although the overall results indicate stable and high predictive performance, the Friedman test did not identify statistically significant differences among the algorithms, χ2 = 10.953, p = 0.204. Therefore, model selection should consider not only predictive accuracy but also computational efficiency, calibration, interpretability, and implementation feasibility. Despite the internal stability of the pipeline and satisfactory holdout performance, the public and cross-sectional nature of the dataset limits causal inference and model transferability. Consequently, external and prospective validation is required before integration into institutional early warning systems.

  • Open Access

    Comparative Evaluation of Ensembles in Type 2 Diabetes Classification: Performance, Explainability, and Interpretability

    by Daniel Andrade-Girón, William Marin-Rodriguez, Americo Peña, Enrique Diaz-Vega, Viviana Vellon-Flores, Timoteo Solano-Armas, Algemiro Muñoz-Vilela

    Type 2 diabetes mellitus (T2DM) is a priority public health issue that requires accurate, efficient, and interpretable predictive tools to support the early identification of at-risk cases. Nevertheless, the comparison of machine learning models in this domain is frequently constrained by heterogeneous protocols, a reliance on overall accuracy, and an inadequate integration of metrics for agreement, calibration, robustness, and interpretability. This study employed a comparative approach to evaluate 10 ensemble and boosting models for the binary classification of T2DM under a unified experimental protocol. The evaluation employed nested cross-validation with 10 external folds and 3 internal folds, a 20% internal holdout, and complementary metrics of performance, discrimination, agreement, calibration, and computational cost. In the context of nested cross-validation, the Random Forest algorithm demonstrated a superior performance, attaining the highest average weighted F1-score (92.51 ± 5.80%), the highest Matthews' correlation coefficient (MCC) (0.840 ± 0.120%), and high discriminative power (weighted area under the receiver operating characteristic curve (ROC-AUC) = 97.46 ± 2.12%). Consequently, it was selected based on a predefined composite criterion. CatBoost achieved the highest weighted ROC-AUC (97.91 ± 2.39%) and area under the precision-recall curve (PR-AUC) (96.60 ± 3.61%), while Extra Trees demonstrated performance that was virtually equivalent to that of the selected model. The Friedman test revealed significant overall differences among models (χ² = 27.898607; p = 0.000992). However, the Nemenyi post-hoc test indicated that the leading models were statistically comparable to one another and that the significant differences were concentrated relative to AdaBoost. In the holdout set, the calibrated Random Forest attained an accuracy of 0.882, a balanced accuracy of 0.880, a weighted F1-score of 0.884, an ROC-AUC of 0.943, a PR-AUC of 0.879, a κ of 0.736, an MCC of 0.739, a Brier score of 0.091, and an expected calibration error of 0.130, with a recall of 0.88 for the positive class. Explainability analyses employing SHapley Additive Explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) techniques have been demonstrated to offer traceability and predictive plausibility, whilst eschewing the assumption of causality. The findings of this study suggest that Random Forest, Extra Trees, and CatBoost represent robust alternatives for the tabular classification of T2DM. However, it is important to note that their clinical application requires multicenter external validation, local recalibration, and prospective evaluation.

  • Open Access

    Predicting Student Stress Using Machine Learning Ensemble Models: A Multi-Criteria Comparison and Complementary Explanatory Analysis

    by Daniel Cristóbal Andrade-Girón, William Joel Marin-Rodriguez, Marcelo Gumercindo Zuñiga-Rojas, Abrahán Cesar Neri-Ayala, Edgar Tito Susanibar-Ramírez, Miguel Angel Aguilar-Luna-Victoria

    Student stress is a significant mental health issue in educational settings; therefore, the development of reliable, calibrated, and interpretable predictive models can help classify observed levels of stress. This study analyzed the public Student Stress Factors dataset (N = 1,100; 20 predictors and 1 target variable) using a supervised pipeline designed to minimize information leakage. This pipeline incorporated stratified partitioning, encapsulated preprocessing, and nested cross-validation applied exclusively to the training subsets. A comparative analysis was conducted on nine ensembles and boosting algorithms designed for tabular data: AdaBoost, Gradient Boosting, Random Forest, Extra Trees, Bagging, Voting, Stacking, XGBoost, and LightGBM. The performance of the model was evaluated using several evaluation metrics, including accuracy, balanced accuracy, F1-weighted, Matthews’ correlation coefficient (MCC), receiver operating characteristic—area under the curve (ROC-AUC) weighted, and Brier score. These metrics were supplemented by the nonparametric Friedman test, which was used for model comparison. The findings indicated that the performance was consistently high. Gradient Boosting achieved the highest average performance in nested cross-validation (accuracy = 89.55 ± 3.16%; F1-weighted = 89.54 ± 3.17%; MCC = 0.845 ± 0.047; ROC-AUC weighted = 98.59 ± 0.92%). XGBoost and LightGBM followed closely behind. In the independent holdout set, the final calibrated model exhibited robust performance metrics (accuracy = 0.8818; F1-weighted = 0.8818; MCC = 0.8237; ROC-AUC weighted = 0.9861). These findings substantiate the model’s stability and generalization capability. However, the Friedman test did not reveal significant differences between the algorithms (χ² = 10.953; p = 0.204). Consequently, the selection of a model should take into account not only its predictive performance but also its computational efficiency, calibration, interpretability, and feasibility of implementation. Despite the pipeline’s internal stability and satisfactory holdout performance, the public and cross-sectional nature of the data imposes limitations on causal inference and model transferability. Consequently, external and prospective validations are imperative prior to its integration into institutional early warning systems.

  • Open Access

    Benchmarking of Ensembles and Meta-Ensembles in the Multiclass Classification of Obesity-Status Classification: Predictive Performance, Calibration and Interpretability

    by Daniel Andrade-Girón, William Marin-Rodriguez, Americo Peña, Elsa Oscuvilca-Tapia, Fredy Bermejo-Sanchez

    Obesity is a major public health concern because of its high prevalence and association with cardiometabolic comorbidities. This study compared nine ensemble and meta-ensemble learning models for multiclass obesity-status classification using the Obesity Dataset, comprising 1610 records, 14 predictors, and four body-weight status classes. To ensure a leakage-aware evaluation, all preprocessing and resampling steps were embedded within the validation workflow. Standardization, one-hot encoding, and RandomOverSampler were applied only within the training folds; SMOTE and no-resampling configurations were retained as configurable alternatives but were not used to generate the reported results. Model performance was assessed using complementary classification, discrimination, agreement, and calibration metrics, including accuracy, balanced accuracy, weighted F1-score, macro F1-score, weighted ROC-AUC, Matthews correlation coefficient, Brier score, and multiclass expected calibration error. Overall, the ensemble models achieved strong discriminative performance, with eight of nine classifiers exceeding 82% accuracy and obtaining weighted ROC-AUC values close to or above 94%. LightGBM showed the strongest mean metric-based profile, with an accuracy of 85.41 ± 2.85%, weighted F1-score of 85.25 ± 2.88%, weighted ROC-AUC of 95.58 ± 1.52%, and MCC of 0.779 ± 0.042. Random Forest and Stacking achieved comparable classification performance, although Stacking presented poorer calibration. The Friedman test detected significant global differences among classifiers, χ2 = 38.7733, p = 0.000005. However, the Nemenyi post hoc test indicated that Stacking, Random Forest, LightGBM, Voting, Gradient Boosting, and Extra Trees belonged to the same high-performance statistical group. Therefore, LightGBM was selected as the final model based on its practical balance of predictive performance, calibration behavior, stability, and implementation feasibility, rather than on unequivocal statistical superiority. On the independent holdout set, LightGBM maintained strong generalization, achieving accuracy = 0.8447, weighted F1-score = 0.8435, MCC = 0.7653, and weighted ROC-AUC = 0.9464. Calibration was moderate, with Brier score = 0.2575 and multiclass ECE = 0.1070, indicating that predicted probabilities should be interpreted cautiously when used to support threshold-based decisions.


  • Open Access

    Benchmarking of Ensembles and Meta‐Ensembles in the Multiclass Classification of Obesity Risk: Predictive Performance, Calibration and Interpretability

    by Daniel Andrade-Girón, William Marin-Rodriguez, Américo Peña, Elsa Oscuvilca-Tapia, Fredy Bermejo-Sanchez

    Obesity represents a significant public health concern, attributable to its high preva-lence and its association with cardiometabolic comorbidities. This study compared a set of ensemble learning models—including canonical ensembles, meta-ensembles, and base-lines for tabular data—in a multiclass obesity status prediction task using the “Obesity Dataset” (n = 1,610; 14 predictors; 4 classes). To ensure methodological rigor, a pipeline was implemented using ColumnTransformer, standardization, one-hot encoding, and re-balancing via SMOTENC applied exclusively to the training folds, thereby preventing data leakage. The performance of the system was evaluated using several evaluation metrics, including accuracy, F1-score, precision, recall, Cohen's kappa, and Matthews correlation coefficient. This evaluation was supplemented by a computational cost analysis. Inferen-tial comparisons were executed using the Friedman test and the Nemenyi post-hoc test (α = 0.05). The findings indicated a high level of overall performance (≈89–90.5% precision), identifying a leading group of models that were statistically indistinguishable (Group A). This group included LightGBM (90.49% ± 1.38), Random Forest (90.16% ± 1.70), Stacking (90.21% ± 1.70), and Extra Trees (89.69% ± 1.55). It has been demonstrated that models such as XGBoost, Bagging, and CatBoost demonstrate competitive performance with par-tial statistical overlap. Conversely, Gradient Boosting and AdaBoost exhibited signifi-cantly lower performance. In summary, a single dominant model was not identified; ra-ther, a set of equivalent solutions was identified. The selection of a model should be based on a balance between accuracy, computational cost, and interpretability. Random Forest and Extra Trees are efficient options, and Stacking is a valid alternative when maximizing predictive performance is prioritized.


  • Open Access

    Comparison of Ensemble and Meta-Ensemble Models for Early Risk Prediction of Acute Myocardial Infarction

    by Daniel Cristóbal Andrade-Girón, Juana Sandivar-Rosas, William Joel Marin-Rodriguez, Marcelo Gumercindo Zúñiga-Rojas, Abrahán Cesar Neri-Ayala, Ernesto Díaz-Ronceros

    Cardiovascular disease (CVD) is a major cause of mortality around the world. This underscores the critical need to implement effective predictive tools to inform clinical decision-making. This study aimed to compare the predictive performance of ensemble learning algorithms, including Bagging, Random Forest, Extra Trees, Gradient Boosting, and AdaBoost, when applied to a clinical dataset comprising patients with CVD. The methodology entailed data preprocessing and cross-validation to regulate generalization. The performance of the model was evaluated using a variety of metrics, including accuracy, F1 score, precision, recall, Cohen’s Kappa, and area under the curve (AUC). Among the models evaluated, Bagging demonstrated the best overall performance (accuracy ± SD: 93.36% ± 0.22; F1 score: 0.936; AUC: 0.9686). It also reached the lowest average rank (1.0) in Friedman test and was placed, together with Extra Trees (accuracy ± SD: 90.76% ± 0.18; F1 score: 0.916; AUC: 0.9689), in the superior statistical group (group A) according to Nemenyi post hoc test. The two models demonstrated a high degree of agreement with the actual labels (Kappa: 0.87 and 0.83, respectively), thereby substantiating their reliability in authentic clinical contexts. The findings substantiated the preeminence of aggregation-based ensemble methods in terms of accuracy, stability, and concordance. This underscored the prominence of Bagging and Extra Trees as optimal candidates for cardiovascular diagnostic support systems, where reliability and generalization were paramount.


  • Open Access

    Comparison of Assembly Methods in Machine Learning for the Early Prediction of Acute Myocardial Infarction

    by Daniel Cristóbal Andrade-Girón, Juana Sandivar-Rosas, William Joel Marin-Rodriguez, Marcelo Gumercindo Zuñiga-Rojas, Abrahán Cesar Neri-Ayala, Ernesto Díaz-Ronceros

    Cardiovascular disease (CVD) is one of the leading causes of mortality worldwide, which raises the fundamental need to apply efficient predictive tools to support clinical deci-sion-making. This study compares the predictive performance of Bagging, Random Forest, Extra Trees, Gradient Boosting, and AdaBoost ensemble learning algorithms applied to a clinical dataset of CVD patients. The methodology included data preprocessing and cross-validation to regulate generalization. Performance was evaluated using multiple metrics: accuracy, F1 score, precision, recall, Cohen's Kappa, and area under the curve (AUC). Among the models evaluated, Bagging demonstrated the best overall performance (accuracy: 93.36 %± 0.22; F1 Score: 0.936; AUC: 0.9686), reaching also the lowest average rank (1.0) in Friedman's test and placing, together with Extra Trees (90.76 %± 0.18; AUC: 0.9689), in the superior statistical group (group A) according to Nemenyi's post-hoc test. Both models evidenced high agreement with the real labels (Kappa: 0.87 and 0.83, respec-tively), which reinforces their reliability in real clinical settings. The results validate the superiority of aggregation-based ensemble methods in terms of accuracy, stability, and concordance, highlighting Bagging and Extra Trees as preferred candidates for cardio-vascular diagnostic support systems, where reliability and generalization are essential.

  • Open Access

    Intelligent Feature Selection Ensemble Model for Price Prediction in Real Estate Markets

    by Daniel Cristóbal Andrade-Girón, William Joel Marin-Rodriguez, Marcelo Gumercindo Zuñiga-Rojas

    Real estate is crucial to the global economy, propelling economic and social development. This study examines the effects of dimensionality reduction through Recursive Feature Elimination (RFE), Random Forest (RF), and Boruta on real estate price prediction, assessing ensemble models like Bagging, Random Forest, Gradient Boosting, AdaBoost, Stacking, Voting, and Extra Trees. The results indicate that the Stacking model achieved the best performance with an MAE (mean absolute error) of 14,090, MSE (mean squared error) of 5.338 × 108, RMSE (root mean square error) of 23,100, R2 of 0.924, and a Concordance Correlation Coefficient (CCC) of 0.960, also demonstrating notable computational efficiency with a time of 67.23 s. Gradient Boosting closely followed, with an MAE of 14,540, R2 of 0.920, and a CCC of 0.958, requiring 1.76 s for computation. Variable reduction through RFE in both Gradient Boosting and Stacking led to an increase in MAE by 16.9% and 14.6%, respectively, along with slight reductions in R2 and CCC. The application of Boruta reduced the variables to 16, maintaining performance in Stacking, with an increase in MAE of 9.8% and a R2 of 0.908. These dimensionality reduction techniques enhanced computational efficiency and proved effective for practical applications without significantly compromising accuracy. Future research should explore automatic hyperparameter optimization and hybrid approaches to improve the adaptability and robustness of models in complex contexts.


  • Open Access

    Intelligent Feature Selection Ensemble Model for Price Prediction in Real Estate Markets

    by Daniel Cristobal Andrade-Girón, William Joel Marin-Rodriguez, Marcelo Gumercindo Zuñiga-Rojas

    Real estate is crucial to the global economy, propelling economic and social development. This study examines the effects of dimensionality reduction through Recursive Feature Elimination (RFE), Random Forest (RF), and Boruta on real estate price prediction, as-sessing ensemble models like Bagging, Random Forest, Gradient Boosting, AdaBoost, Stacking, Voting, and Extra Trees. The findings indicate that the Stacking model achieved the best performance with an MAE of 14,092.30, an MSE of 5.34 × 10⁸, an RMSE of 23,104.19, and an R² of 0.9241, followed closely by Gradient Boosting (MAE = 14,536.52, R² = 0.9197). However, applying RFE for a variable reduction in Gradient Boosting led to a 16.9% increase in MAE and a 1.6% decrease in R². A similar trend was noted in Stacking, where the complete version had a 14.6% lower MAE. RF also displayed variable impacts: Gradient Boosting and Stacking saw MAE increases of 19.1% and 17.7%, respectively, whereas RF combined with AdaBoost enhanced performance with a 5.3% reduction in MAE. Boruta enabled the reduction of variables to 16 without significantly impacting ac-curacy; in Stacking, the MAE only increased by 9.8%, while R² decreased to 0.9082. Beyond accuracy, dimensionality reduction enhances computational efficiency, promoting scala-bility in real applications. Future research should investigate hyperparameter optimiza-tion and hybrid strategies to boost performance in complex settings.