Chemometric and computational modeling of polysaccharide coated drugs for colonic drug delivery DOI Creative Commons
Ahmed Al‐Omari, Khaled Almansour

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 26, 2025

A methodology based on Principal Component Analysis (PCA) and machine learning (ML) regression was developed in this study for predicting 5-aminosalicylic acid drug release from polysaccharide-coated formulation. The Raman method used collection of spectral data which were then as inputs to the ML models estimation release. For modeling, we examined predictive accuracy three models-Elastic Net (EN), Group Ridge Regression (GRR), Multilayer Perceptron (MLP)-for forecasting behavior samples. dataset, consisting 155 points with over 1500 features, underwent preprocessing involving normalization, dimensionality reduction, outlier detection using Cook's Distance. Model hyperparameters tuned Slime Mould Algorithm (SMA), each model's performance evaluated through k-fold cross-validation (k = 3). Assessment metrics, such coefficient determination (R²), root mean square error (RMSE), absolute (MAE), emphasize MLP exceptional performance. On test set, achieved an R² 0.9989, notably higher than EN's 0.9760 GRR's 0.7137. Additionally, exhibited remarkably low RMSE MAE values at 0.0084 0.0067, respectively, comparison 0.0342 0.0267, well 0.0907 0.0744. Parity plots curves further validate MLP's reliability, demonstrating close alignment between actual predicted efficient minimal overfitting. Consequently, model emerges most effective approach task, offering a robust tool accurately modeling complex data. These findings underscore robustness model, providing reliable formulations, implications advancing colonic delivery systems.

Язык: Английский

Dendritic Growth Optimization: A Novel Nature-Inspired Algorithm for Real-World Optimization Problems DOI Creative Commons
Ishaani Priyadarshini

Biomimetics, Год журнала: 2024, Номер 9(3), С. 130 - 130

Опубликована: Фев. 21, 2024

In numerous scientific disciplines and practical applications, addressing optimization challenges is a common imperative. Nature-inspired algorithms represent highly valuable pragmatic approach to tackling these complexities. This paper introduces Dendritic Growth Optimization (DGO), novel algorithm inspired by natural branching patterns. DGO offers solution for intricate problems demonstrates its efficiency in exploring diverse spaces. The has been extensively tested with suite of machine learning algorithms, deep metaheuristic the results, both before after optimization, unequivocally support proposed algorithm’s feasibility, effectiveness, generalizability. Through empirical validation using established datasets like diabetes breast cancer, consistently enhances model performance across various domains. Beyond working experimental analysis, DGO’s wide-ranging applications learning, logistics, engineering solving real-world have highlighted. study also considers implications implementing multiple scenarios. As remains crucial research industry, emerges as promising avenue innovation problem solving.

Язык: Английский

Процитировано

4
Chemometric and computational modeling of polysaccharide coated drugs for colonic drug delivery DOI Creative Commons
Ahmed Al‐Omari, Khaled Almansour

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 26, 2025

A methodology based on Principal Component Analysis (PCA) and machine learning (ML) regression was developed in this study for predicting 5-aminosalicylic acid drug release from polysaccharide-coated formulation. The Raman method used collection of spectral data which were then as inputs to the ML models estimation release. For modeling, we examined predictive accuracy three models-Elastic Net (EN), Group Ridge Regression (GRR), Multilayer Perceptron (MLP)-for forecasting behavior samples. dataset, consisting 155 points with over 1500 features, underwent preprocessing involving normalization, dimensionality reduction, outlier detection using Cook's Distance. Model hyperparameters tuned Slime Mould Algorithm (SMA), each model's performance evaluated through k-fold cross-validation (k = 3). Assessment metrics, such coefficient determination (R²), root mean square error (RMSE), absolute (MAE), emphasize MLP exceptional performance. On test set, achieved an R² 0.9989, notably higher than EN's 0.9760 GRR's 0.7137. Additionally, exhibited remarkably low RMSE MAE values at 0.0084 0.0067, respectively, comparison 0.0342 0.0267, well 0.0907 0.0744. Parity plots curves further validate MLP's reliability, demonstrating close alignment between actual predicted efficient minimal overfitting. Consequently, model emerges most effective approach task, offering a robust tool accurately modeling complex data. These findings underscore robustness model, providing reliable formulations, implications advancing colonic delivery systems.

Язык: Английский

Процитировано

0