Dimensionality reduction, and function approximation of poly(lactic-co-glycolic acid) micro- and nanoparticle...

Area of Science:

• Materials Science
• Pharmaceutical Sciences
• Computational Chemistry

Background:

• Accurate prediction of poly(lactic-co-glycolic acid) (PLGA) micro- and nanoparticle dissolution rates is critical for drug development and manufacturing.
• The complexity arising from a large number of features (300) and high redundancy in PLGA dissolution datasets poses significant challenges for accurate predictive modeling.

Purpose of the Study:

• To develop a more accurate model for predicting PLGA dissolution rates.
• To address the challenges of high dimensionality and feature redundancy in PLGA dissolution prediction.

Main Methods:

• Application of dimensionality reduction techniques to simplify the feature space.
• Independent evaluation of a diverse set of regression algorithms.
• Development and testing of ensemble methods to enhance prediction accuracy.

Main Results:

• Dimensionality reduction techniques were employed to manage the large feature set.
• Various regression and ensemble methods were assessed for predictive performance.
• The proposed evolutionary weighted ensemble method demonstrated the lowest margin of error.

Conclusions:

• The evolutionary weighted ensemble method significantly outperformed individual algorithms and other ensemble techniques in predicting PLGA dissolution rates.
• The study highlights the effectiveness of dimensionality reduction and ensemble methods for complex pharmaceutical material property prediction.