• Corpus ID: 222090831

Persistent homology advances interpretable machine learning for nanoporous materials

@article{Krishnapriyan2020PersistentHA,
  title={Persistent homology advances interpretable machine learning for nanoporous materials},
  author={Aditi S. Krishnapriyan and Joseph H. Montoya and Jens S. Hummelsh{\o}j and Dmitriy Morozov},
  journal={ArXiv},
  year={2020},
  volume={abs/2010.00532}
}
Machine learning for nanoporous materials design and discovery has emerged as a promising alternative to more time-consuming experiments and simulations. The challenge with this approach is the selection of features that enable universal and interpretable materials representations across multiple prediction tasks. We use persistent homology to construct holistic representations of the materials structure. We show that these representations can also be augmented with other generic features such… 

Figures and Tables from this paper

PersGNN: Applying Topological Data Analysis and Geometric Deep Learning to Structure-Based Protein Function Prediction
TLDR
It is demonstrated that the hybridized approach, PersGNN, outperforms either method on its own as well as a baseline neural network that learns from the same information, indicating the transferability of this approach.
Topology Applied to Machine Learning: From Global to Local
TLDR
The meta-hypothesis is that the short bars are as important as the long bars for many machine learning tasks and work connecting persistent homology to geometric features of spaces, including curvature and fractal dimension is surveyed.
Learning from learning machines: a new generation of AI technology to meet the needs of science
TLDR
If the fundamental challenges associated with “bridging the gap” between domain-driven scientific models and data-driven AI learning machines are addressed, then it is expected that these AI models can transform hypothesis generation, scientific discovery, and the scientific process itself.
New evidence from exceptionally “well-preserved” specimens sheds light on the structure of the ammonite brachial crown
TLDR
It appears that there are at least three independent evolutionary origins of hooks: in belemnoids, oegospids, and now in ammonites, and for the first time a hypothetical reconstruction of an ammonite brachial crown is proposed.

References

SHOWING 1-10 OF 36 REFERENCES
Topological Descriptors Help Predict Guest Adsorption in Nanoporous Materials
TLDR
By investigation of the importance of individual topological features in the adsorption model, this work is able to pinpoint the location of the pores that correlate best to adsor adaptation at different pressure, contributing to the atom-level understanding of structure-property relationships.
Big-Data Science in Porous Materials: Materials Genomics and Machine Learning
TLDR
It is shown that having so many materials allows us to use big-data methods as a powerful technique to study these materials and to discover complex correlations.
Chemically intuited, large-scale screening of MOFs by machine learning techniques
A novel computational methodology for large-scale screening of MOFs is applied to gas storage with the use of machine learning technologies. This approach is a promising trade-off between the
Role of Pore Chemistry and Topology in the CO2 Capture Capabilities of MOFs: From Molecular Simulation to Machine Learning
Open framework materials (OFMs) such as metal–organic frameworks (MOFs) can provide structurally and chemically tailorable nanopores. This exceptional tunability has allowed for careful positioning
Applications of machine learning in metal-organic frameworks
A Universal Machine Learning Algorithm for Large Scale Screening of Materials.
TLDR
This work introduces chemical intuition in the authors' descriptors by using the "type" of the atoms in the structure, instead of the previously used building blocks, to account for the chemical character of the MOF.
Understanding the diversity of the metal-organic framework ecosystem
TLDR
A machine learning method is developed to quantify similarities of MOFs to analyse their chemical diversity and identifies biases in the databases, and it is shown that such bias can lead to incorrect conclusions.
Quantifying similarity of pore-geometry in nanoporous materials
TLDR
A pore recognition approach is developed to quantify similarity of pore structures and classify them using topological data analysis, which allows to identify materials with similar pore geometries, and to screen for materials that are similar to given top-performing structures.
Geometrical Properties Can Predict CO2 and N2 Adsorption Performance of Metal-Organic Frameworks (MOFs) at Low Pressure.
TLDR
K-means clustering and archetypal analysis are applied to identify the truly significant nanoporous structures in a large library of ∼82 000 virtual MOFs to serve as efficient filtering tools to detecting promising high-performing candidates at the early stage of virtual high-throughput screening of novel porous materials.
Persistence Images: A Stable Vector Representation of Persistent Homology
TLDR
This work converts a PD to a finite-dimensional vector representation which it is called a persistence image, and proves the stability of this transformation with respect to small perturbations in the inputs.
...
...