Multi-modal cascade feature transfer for polymer property prediction

2025-05-06

AI Summary

Existing polymer property prediction methods relying on single-modality modeling—using only chemical structures, molecular descriptors, or additive information—are limited in representational capacity. To address this, we propose a cascaded multimodal feature transfer framework that establishes, for the first time, an end-to-end co-learning mechanism integrating graph convolutional network (GCN)-driven structural representations with molecular descriptors and additive features. Our approach enables complementary modeling of heterogeneous data through cross-modal feature alignment, cascaded fusion, and joint optimization. Evaluated across multiple polymer benchmark datasets, the framework achieves an average 18.7% reduction in prediction error and demonstrates significantly improved generalization performance for critical properties—including glass transition temperature and tensile strength. This work introduces a novel paradigm for fusing multi-source data in polymer informatics, advancing the state of multimodal representation learning for functional material design.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

In this paper, we propose a novel transfer learning approach called multi-modal cascade model with feature transfer for polymer property prediction.Polymers are characterized by a composite of data in several different formats, including molecular descriptors and additive information as well as chemical structures. However, in conventional approaches, prediction models were often constructed using each type of data separately. Our model enables more accurate prediction of physical properties for polymers by combining features extracted from the chemical structure by graph convolutional neural networks (GCN) with features such as molecular descriptors and additive information. The predictive performance of the proposed method is empirically evaluated using several polymer datasets. We report that the proposed method shows high predictive performance compared to the baseline conventional approach using a single feature.

Problem

• Predicts polymer properties using multi-modal data fusion
• Combines GCN features with molecular descriptors
• Improves accuracy over single-feature baseline methods

Innovation

• Multi-modal cascade model with feature transfer
• Combines GCN features with molecular descriptors
• Empirically validated on multiple polymer datasets