Unsupervised Multi-Clustering and Decision-Making Strategies for 4D-STEM Orientation Mapping
This work addresses the challenge of reliable orientation mapping for researchers in materials science, though it appears incremental as it builds on existing methods like NMF with new preprocessing and evaluation techniques.
The study tackled the problem of determining the optimal number of components for robust orientation mapping in 4D-STEM datasets by integrating unsupervised learning with decision-making strategies, resulting in a framework that balances reconstruction fidelity and model complexity using NMF, K-Component Loss, and IQA metrics.
This study presents a novel integration of unsupervised learning and decision-making strategies for the advanced analysis of 4D-STEM datasets, with a focus on non-negative matrix factorization (NMF) as the primary clustering method. Our approach introduces a systematic framework to determine the optimal number of components (k) required for robust and interpretable orientation mapping. By leveraging the K-Component Loss method and Image Quality Assessment (IQA) metrics, we effectively balance reconstruction fidelity and model complexity. Additionally, we highlight the critical role of dataset preprocessing in improving clustering stability and accuracy. Furthermore, our spatial weight matrix analysis provides insights into overlapping regions within the dataset by employing threshold-based visualization, facilitating a detailed understanding of cluster interactions. The results demonstrate the potential of combining NMF with advanced IQA metrics and preprocessing techniques for reliable orientation mapping and structural analysis in 4D-STEM datasets, paving the way for future applications in multi-dimensional material characterization.