Scalable Machine Learning Models for Predicting Quantum Transport in Disordered 2D Hexagonal Materials

TL;DR

This paper develops scalable machine learning models to predict quantum transport properties in disordered 2D hexagonal materials, demonstrating their accuracy and limitations across diverse configurations and highlighting the need for physics-informed approaches.

Contribution

It introduces a geometry-driven, interpretable feature space and evaluates random forest models for predicting quantum transport in 2D materials, addressing generalization challenges.

Findings

Regression models outperform classification in in-domain predictions.

Model extrapolation to unseen regimes is significantly limited.

A physically interpretable feature space enhances model generalization.

Abstract

We introduce scalable machine learning models to accurately predict two key quantum transport properties, the transmission coefficient T(E) and average local density of states (Average-LDOS) in two-dimensional (2D) hexagonal materials with magnetic disorder. Using a tight binding Hamiltonian combined with the Non-Equilibrium Green's Function (NEGF) formalism, we generate a large dataset of over 400,000 unique configurations across graphene, germanene, silicene, and stanene nanoribbons with varying geometries, impurity concentrations, and energy levels. A central contribution of this work is the development of a geometrydriven, physically interpretable feature space that enables the models to generalize across material types and device sizes. Random Forest regression and classification models are evaluated in terms of accuracy, stability, and extrapolation ability. Regression consistently outperforms classification in capturing continuous transport behavior on in-domain data. However, extrapolation performance degrades significantly, revealing the limitations of tree-based models in unseen regimes. This study highlights both the potential and constraints of scalable ML models for quantum transport prediction and motivates future research into physics-informed or graph-based learning architectures for improved generalization in spintronic and nanoelectronic device design.