Four keynote speakers, 26 speakers and tutors from the NOMAD team, and 34 participants made NOMAD Summer 2018 a resounding success. Held at the CECAM Headquarters (Lausanne, 24 - 27 September 2018) overlooking Lake Geneva, the school introduced researchers from academia and industry to state-of-the-art methods and practical tools in Big-Data-driven materials science developed within NOMAD. The program included keynote addresses, lectures and hands-on tutorials, but also allowed for numerous and interesting discussions with participants, including representatives of world-leading companies. The greatest interest was in the application of the various tools developed by NOMAD to industrially-relevant problems and how quality can be assured, as well as scientific reproducibility.

If you weren't able to join us, a playlist of the lectures is available on YouTube!

More information on the tutorials from the course is also available here.

Welcome Video

Data Repositories, Data Archives and Metadata

M. Scheffler: A primer to NOMAD Video

KEYNOTE, J. Vreeken: What is Going On in My Data? How to Discover Interpretable Correlations, Associations, and Causation Video

F. Mohamed, L. Ghiringhelli: Hands-on session - Upload, search and retrieve materials science data with the NOMAD Repository and the NOMAD Archive  Video Exercise 1 Video Exercise 2

NOMAD Encyclopedia

M. De Mier: Technology Transfer: from research to industry Video

C. Draxl:  Making scientific data accessible: the NOMAD Encyclopedia Video

G. Huhs: Hands-on session - The NOMAD Encyclopedia 

Data Analytics I: Overview, Infrastructure and Query

KEYNOTE, A. Chandrasekaran: Polymer genome: An informatics platform for rational polymer dielectrics design and beyond Video

L. Ghiringhelli The NOMAD analytics toolkit: interactive big-data driven materials science over the web 

R. Ouyang SISSO: a compressed-sensing method for systematically identifying efficient physical models of materials properties Video

L. Ghiringhelli, A. Ziletti, E. Ahmetcik, B. Hoock: Hands-on session - Data-driven materials science with the NOMAD Analytics Toolkit Video Exercise 1 Video Exercise 2

Kaggle Competition

C. Sutton: Introduction to the NOMAD Kaggle competition 2018: Predicting Transparent Conductors

Kaggle competition 1st place winner, T. Yamamoto: The n-grams approach and it application to materials science Video

KEYNOTE, Kaggle competition 2nd place winner, Y. Lysogorskiy: Locally approximated electronic structure based descriptors for predicting materials properties Video

KEYNOTE, Kaggle competition 3rd place winner, L. Blumenthal: Building neural networks on top of smooth overlap of atomic position (SOAP) descriptors

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

A. Ziletti: Insightful crystal-structure classification using deep learning

M. Boley: Subgroup discovery in materials science Video

S. Rigamonti, M. Troppenz: Cluster expansion Video Video

A. Ziletti, M. Boley, C. Sutton, S. Rigamonti, M. Troppenz: Hands-on session - Neural networks, subgroup discovery and cluster expansion within the NOMAD Analytics Toolkit

High-throughput Calculations & Data Quality

G. Hautier, C. Carbogno: High-throughput calculations, error estimates in density functional theory Video Video

G. Hautier: Hands-on session - Predicting density functional theory error estimates with the NOMAD Analytics Toolkit

NOMAD Advanced Graphics

KEYNOTE, U. Woessner: Virtual and Augmented Reality for Scientific Visualization Video

R. Garcia, U. Woessner, M. Compostella, M. Rampp: Hands-on session - Introduction to the NOMAD Advanced Graphics Toolkit, and virtual-reality demonstrations

Data Analytics III: Kernel Methods for supervised and unsupervised learning

G. Csanyi: Overview of kernel-based machine learning of atomistic properties Video

KEYNOTE, M. Ceriotti: Machine-learning of materials between predictions and insights Video

A. Fekete, M. Stella, G. Csanyi: Hands-on session - The SOAP descriptor, Gaussian Approximation Potentials (GAP) and machine-learning of force fields within the NOMAD Analytics Toolkit