Oct 24, 2017 · In this paper, we propose an atomic distance kernel, which can incorporate the crystal structure information into the prediction without ...

In this research, we propose a novel kernel function between compounds, which directly uses crystal structure information for the prediction of physical ...

In this research, we propose a novel kernel function between compounds, which directly uses crystal structure information for the prediction of physical ...

Oct 2, 2023 · This paper presents a study of various WL kernel models and ECFP-based learning models for predicting the electron affinity, ionization ...

Dec 22, 2023 · Here we propose a set of five realistic materials property prediction benchmark problems, in which the test samples are located in sparse composition or ...

May 2, 2024 · This encoding is done at the pair level, using the Euclidean distance between atoms and a pair-type aware Gaussian kernel. A transformer ...

To apply the marginalized graph kernel, a spatial adjacency rule is first applied to convert molecules into graphs whose vertices and edges are labeled by ...

(60,61) The vectors generated by the AABBA kernel were leveraged in ML models predicting TMC properties, including neural networks (NNs), gradient boosting ...

Oct 2, 2023 · The results demonstrate that the GPR/WL kernel methods can accurately predict the electronic properties of PAHs and their derivatives with root ...

Oct 16, 2023 · Here, we present a general class of material descriptors motivated by the machine-learning theory of kernel mean embedding.