AI-Based Polymer Classification Using Ensemble Deep Learning and Heuristic Optimization: Implications for Recycling...

Area of Science:

• Materials Science
• Computer Science
• Environmental Science

Background:

• Increasing global polymer production necessitates sustainable solutions for consumption, production, and disposal.
• Biodegradable and bio-based polymers offer promising avenues to mitigate plastic pollution and resource depletion.
• Effective polymer classification is crucial for advancing materials research and recycling strategies.

Purpose of the Study:

• To develop and evaluate a novel ensemble deep learning model for accurate polymer-type classification.
• To investigate the impact of feature selection using heuristic optimization on classification performance.
• To assess the model's effectiveness under limited and unbalanced data conditions.

Main Methods:

• The study introduces the New Polymers Frontier in Recycling and Sustainability Using an Ensemble of Deep Learning with a Heuristic Search Algorithm (NPFRS-EDLHSA).
• The NPFRS-EDLHSA model integrates bidirectional recurrent neural network (BiRNN), bidirectional gated recurrent unit (BiGRU), and graph autoencoder (GAE) techniques.
• Hyperparameter optimization is performed using the grasshopper optimization algorithm (GOA) for enhanced classification.

Main Results:

• The NPFRS-EDLHSA model demonstrated superior performance in polymer-type classification compared to existing methods.
• Feature selection with heuristic optimization positively influenced classification accuracy, particularly under data limitations.
• The ensemble deep learning approach proved effective for classifying polymers on small, unbalanced datasets.

Conclusions:

• The NPFRS-EDLHSA model provides a robust methodological framework for computational polymer typology.
• This approach facilitates downstream materials research by improving polymer classification accuracy.
• The findings highlight the potential of advanced machine learning techniques in addressing challenges in polymer sustainability and recycling.