NOMAD Summer A hands-on course on tools for novel-materials discovery

NOMAD Summer

A hands-on course on tools for novel-materials discovery
September 24 - 27, 2018, Lausanne, Switzerland

NOMAD summer tutorials

Note

Since most hands-on sessions will be web-based, we advise the participants to make an account on NOMAD to access the services within a personal workspace, and keep their work after the NOMAD school. To create an account go to https://analytics-toolkit.nomad-coe.eu/ and click on the green “Login” in the gray bar.

1. NOMAD Encyclopedia

With this tutorial you will familiarize with the NOMAD Encyclopedia, its application program interface (API), and the custom-made 3D viewer.


2. Data Analytics I: Overview, Infrastructure, and Query

Predicting energy differences between crystal structures: (Meta-)stability of octet-binary compounds
This tutorial shows how to find descriptive parameters (short formulas) that predict the crystal structure (here, rocksalt (RS), zincblende (ZB), CsCl, NiAs or CrB), using the example of octet binary compounds. It is based on the algorithm Sure Independence Screening followed by L0 minimization (SIS+L0), that enables to search for optimal descriptor by scanning huge feature spaces.
Link to tutorial

Hands-on Workshop Density-Functional Theory and Beyond: Compressed sensing for identifying descriptors
Fully fledged tutorial for understanding both the methodology and the data stream of the first tutorial.
Link to tutorial


3. Data Analytics II: Compress sensing, Cluster Expansion, Neural Networks

Discovering simple descriptors for crystal-structure classification
Subgroup discovery (SGD) is a data mining technique that allows for the discovery of structure in data. It reveals subsets in a given dataset that have unusual or exceptional properties and characterizes such revealed subsets with a natural-language style description.
Link to tutorial

Predicting ground-states of binary alloys through cluster expansion
In this tutorial, the ground state configurations of a SiGe binary alloy are found. Starting from a set of ab initio data for random configurations of the alloy, a cluster expansion is performed and the ground states are found through a configurational sampling.
Link to tutorial


4. High-throughput Calculations & Data Quality

Error estimates from high-accuracy electronic structure reference calculations
A set of tools to analyze the error in electronic structure calculations due to the choice of numerical settings. We use the NOMAD infrastructure to systematically investigate the deviances in total and relative energies as function of typical settings for basis sets, k-grids, etc. for 71 elemental and 81 binary solids in four different electronic-structure codes.
Link to tutorial


5. Advanced Graphics

A guided tutorial on remote-visualization of NOMAD datasets

Interactive exploration of NOMAD datasets in virtual reality (HTC Vive)

  • a guided tutorial on data preparation of the molecular dynamics simulation "Thermal equilibration of Spironapthopyran" for NOMAD VR is available here
  • the example dataset can be
  • a general overview of NOMAD VR for HTC Vive is available here

6. Data Analytics III: Kernel Methods for supervised and unsupervised learning

Learning atomic charges
In this tutorial, we will use Gaussian process regression, GPR (or equivalently, Kernel Ridge Regression, KRR) to train and predict charges of atoms in small organic molecules.
Link to tutorial

Classification of grain boundaries in Fe
In this tutorial we will be using a machine learning method (clustering) to analyse results of Grain Boundary (GB) calculations of alpha-iron.
Link to tutorial

Complexity tool estimator for accurate forces learning: Monocrystalline Silicon example
The goal of this tutorial is to show a possible application of kernel methods for forces prediction using NOMAD data and NOMAD infrastructure. This example is for classical MD data.
To access the tutorial:

[+]     

Page last modified on September 27, 2018, at 11:51 AM EST