# Interpretable Multimodal Machine Learning Analysis of X-ray Absorption Near-Edge Spectra and Pair Distribution Functions

arXiv: https://arxiv.org/abs/2410.17467 · HTML: https://mlsanity.com/paper/2410.17467/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 76 references that are in the map.
- First posted on arXiv 22 Oct 2024.

## Builds on (most cited first, 4 of 4)

- [Multimodal Machine Learning: A Survey and Taxonomy](https://mlsanity.com/paper/1705.09406/) (2017)
- [Spectroscopy-Guided Discovery of Three-Dimensional Structures of Disordered Materials with Diffusion Models](https://mlsanity.com/paper/2312.05472/) (2023)
- [Decoding Structure-Spectrum Relationships with Physically Organized Latent Spaces](https://mlsanity.com/paper/2301.04724/) (2023)
- [Towards End-to-End Structure Solutions from Information-Compromised Diffraction Data via Generative Deep Learning](https://mlsanity.com/paper/2312.15136/) (2023)

## Built on by (most cited first, 4 of 4)

- [XANE(3): An E(3)-Equivariant Graph Neural Network for Accurate Prediction of XANES Spectra from Atomic Structures](https://mlsanity.com/paper/2604.12140/) (2026)
- [Representation of Inorganic Synthesis Reactions and Prediction: Graphical Framework and Datasets](https://mlsanity.com/paper/2512.02947/) (2025)
- [InSpecLearn4SDL: Interpretable Spectral Features Predict Conductivity in Self-Driving Doped Conjugated Polymer Labs](https://mlsanity.com/paper/2509.21330/) (2025)
- [Machine Learning-Driven Crystal System Prediction for Perovskites Using Augmented X-ray Diffraction Data](https://mlsanity.com/paper/2602.04435/) (2026)

## Abstract (arXiv)

We used interpretable machine learning to combine information from multiple heterogeneous spectra: X-ray absorption near-edge spectra (XANES) and atomic pair distribution functions (PDFs) to extract local structural and chemical environments of transition metal cations in oxides. Random forest models were trained on simulated XANES, PDF, and both combined to extract oxidation state, coordination number, and mean nearest-neighbor bond length. XANES-only models generally outperformed PDF-only models, even for structural tasks, although using the metal's differential PDFs (dPDFs) instead of total PDFs narrowed this gap. When combined with PDFs, information from XANES often dominates the prediction. Our results demonstrate that XANES contain rich structural information and highlight the utility of species-specificity. This interpretable, multimodal approach is quick to implement with suitable databases and offers valuable insights into the relative strengths of different modalities, guiding researchers in experiment design and identifying when combining complementary techniques adds meaningful information to a scientific investigation.