DRAFT

In the Framework of the 21^st International Conference “CODATA 2008”

SCIENTIFIC INFORMATION FOR SOCIETY:

FROM TODAY TO THE FUTURE

"Kyiv School": Data Science and Design Science on Materials

Organized by

The University of Tokyo

NSF International Materials Institute

Combinatorial Sciences and Materials Informatics Collaboratory (CoSMIC)

Ukrainian-Japanese Centre

and Metal Physics Department, Physical Engineering Faculty,

of the National Technical University of Ukraine "Kyiv Polytechnic Institute"

Sponsored by

The University of Tokyo

NSF international Materials Institute

Combinatorial Sciences and Materials Informatics Collaboratory (CoSMIC)

3-4 October, 2008, Kyiv, Ukraine

Venue: Ukrainian-Japanese Centre of the National Technical University of Ukraine

“Kyiv Polytechnic Institute”

General Information

From “Kyiv School” Masters: Prof. Shuichi Iwata, Prof. Krishna Rajan

and Prof. Sergey Sidorenko.

_______________________________________________________________

Prof. Shuichi Iwata:

This international “Kyiv School” on “Data Science and Design Science on Materials” is designed for students and young scientists to share pleasures and missions in materials sciences and engineering with world leading scientists who explored new dimensions of data science and design science on materials.

Materials science seeks to understand structure–property relationships, which may be very complex and difficult to discover.

We look for patterns in data that have multiple and different length and time scales. It is rarely possible to construct a single multiscale theory or experiment which can meaningfully and accurately capture relevant information. Data driven design will permit us to survey complex, multiscale information in a high throughput, statistically robust, and yet physically meaningful manner. The application of such approaches can have a significant impact in materials design and discovery.

As this approach to materials science is still in its infancy, it is important to introduce these concepts to the new generation of scientists entering materials science and engineering. They will need to learn an interdisciplinary field that merges ideas from computational materials science and materials theory with machine learning, databases, and parallel and distributed computing and combinatorial experimentation that can integrate data with scientific principles to enable materials discovery and enhance our understanding of structure property relations.

As the fields on materials science and engineering are huge and containing rich legacies reflecting the history of human beings, younger generations need to take advantage of these new advances in materials science and this summer school is part of a global effort in integrating research with education.

________________________________________________________________

Prof. Rajan Krishna:

The aim of this workshop is to examine design of materials as related to electronic structure, crystal structure and thermodynamics from the perspective of data science. The lectures will explore the challenges of acquiring data via experimental or computation means as well as the analysis of that data through coupling it to theory and informatics methods. The applications of these approaches will cover a broad range of materials including intermetallics, crystalline ceramics, nanostructures and molten salts to mention a few. The workshop will also have a component that will be held as a joint meeting with the CODATA Task Group on Materials Data Exchange and Operability. This will be used to examine data handling practices in the materials science community.

________________________________________________________________

Prof. Sergey Sidorenko:

Documents and results of the workshop will be published by National Technical University of Ukraine “Kyiv Polytechnic Institute” and disseminated also internationally through web based modules and notes, Webcasts and other publications.

We'd like to propose to combine “Kyiv School” with the Third NTUU “KPI” International Students Conference “To High Technologies on the Base of Advanced Physical Materials Science Research and Computer Modeling” on the memory of the 100^th anniversary birthday of Academician V.N.Grydnev – Metals Physics Department Founder, Rector of KPI in 1952-1955.

Venue: Ukrainian-Japanese Centre and Metal Physics Department, Physical Engineering Faculty of the National Technical University of Ukraine “Kyiv Polytechnic Institute”.

_______________________________________________________________

٭ ٭ ٭ ٭ ٭

Draft

“Kyiv School” Programme School Masters: Prof. Shuichi Iwata, Prof. Krishna Rajan and Prof. Sergey Sidorenko
September 30, October 1, October 2	Arrivals. Transfer to Hotels and NTUU "KPI" Hostel #19.
October 3
8^30 - 9^00	Registration. Coffee
Ukrainian-Japanese Centre of the NTUU "KPI", Library, 3 floor:
9^00 - 9^10	Greetings by: Michael Zgurovskiy, Rector of the NTUU "KPI", Vladislav Luk'yanov, Deputy of the Verkhovna Rada of Ukraine
9^10 - 10^10	Perspectives on data science and design science on materials by Krishan LAL and Shuichi IWATA
10^10 - 10^30	Coffee Break
10^30 - 11^30	From the first principles calculations to the molecular dynamics for the materials design in terms of the data science by Shuhei OHNISHI
11^30 - 12^30	Designing structure maps: addressing the “curse of dimensionality” problem by Krishna RAJAN
12^30 - 13^30 Catering Halls	Lunch
13^30 - 14^30	Walking around University and the State Polytechnic Museum of Ukraine at NTUU "KPI"
14^30 - 15^30 Ukrainian-Japanese Centre of the NTUU "KPI", Library, 3 floor	The nanostructure problem: Solving the inverse problem for nanostructure from scattering data by Simon BILLINGE
15^30 - 16^30	Data and modeling driven approach towards discovery in materials design by Ying CHEN
Library, Hall #12, 5^th floor:
16^30 - 17^00 in the Hall #12 foyer	Coffee Break
17^00 - 17^30	Fundamentals in Nuclear Fuel Performance by Motoyasu KINOSHITA
17^30 - 18^30	Frontiers of modern materials. Metallic foams – by Sergei SIDORENKO and Olexandra Byakova
18^30 - 20^00	Group discussions with snacks and drink
20^00 - 20^30 in Hall #12	Concert (folk music)
October 4
Ukrainian-Japanese Centre of the NTUU "KPI", Library, 3 floor:
Studies
9^00 - 12^30	Frontiers of data sciences on materials by Toshihiro ASHINO and Laura BARTOLO (Joint Session with CODATA TG Meeting)
12^30 - 14^00	Walking around university (University Laboratory)
14^00 - 15^00	Lunch
15^00 - 17^00	Students reports and discussions: Challenges on data science and design Science on materials after 2007 Summer School by KPI/UT/TU students
17^00	Closing ceremony with Buffet

Abstracts

Perspectives on data science and design science on materials

Shuichi Iwata

The University of Tokyo

Phase changes from data collection and selection into data-driven design and planning are going to happen in many fields. Data quality matters every time and everywhere: gold in, gold out. Reasonable estimations of data uncertainties produce better results and outcomes. The more the problem to be solved is uncertain, the more we should become flexible. Evidence-based deterministic approaches do not work effectively, and adaptive and heuristic approaches work better coupled with in situ data capture, evaluation, and quick decision and timely actions. Holistic creativity as a group is a key for success of the group, when practical maintenance of data quality for proper decision is important.

The time constants of data life cycle are becoming shorter, and the diversity of stakeholders and complexities of data are increasing. New disciplines are continuously created by taking advantage of available data and devices so as to prepare solutions in a timely fashion. Without proper management of continuously-produced important data and without the productivity of new disciplines based on data, we cannot solve important problems of the world.

Materials design was tried by Michael Faraday in 19^th century, and now we have databases, HPC and the Internet. What can we now on materials taking advantage of technologies to manipulate DOS, atomistic structures and microstructures? It’s time enjoy freedom in a decent way.

From the first principles calculations to the molecular dynamics for the materials design in terms of the data science

Shuhei Ohnishi

Toho University, Japan

Discussions are focused on the method how to create effective interatomic potentials from the first principles calculations in computational materials science. Several examples such as d-electron metals and metal/hydrogen systems of clusters, surfaces, and bulk crystals are introduced by the first principles calculations based on the density functional theory. Applying the force data determined by the force-matching method using the first principles calculations, we can investigate dynamical properties not only the equilibrium but also non-equilibrium states by introducing various external environmental parameters like temperature and pressure by the symplectic method in molecular dynamical calculations.

The importance of database by theoretical experimental results with the aids of computers will be discussed for our future work.

Designing structure maps: addressing the “curse of dimensionality” problem

Krishna Rajan

Iowa State University

Structure maps have a long and distinguished history and have formed the foundation in linking crystal chemistry with crystal structure. The challenge in all these maps is that they have a priori assumption as to the important parameters that govern their classification behavior. In this presentation, we show how data mining techniques can help to design new structure maps as well as reproduce established ones. The use of structure maps as design tools for computational and experimental studies for crystallography.

The nanostructure problem: Solving the inverse problem for nanostructure from scattering data

Simon Billinge

Columbia University

Crystallographic methods are the gold standard for atomic structure determination, however a broad and growing class of materials and/or nanophase morphologies do not yield to a crystallographic analysis. The scattering is diffuse and Bragg-peaks become broad and overlapped. This is "the nanostructure problem" which currently has no robust solution. I will discuss our approaches to solving this inverse problem, highlight some of successes and challenges as well as point to a new paradigm of complex or multimodal modeling that is emerging for these problems.

Data and modeling driven approach towards discovery in materials design

Ying Chen

The University of Tokyo

An integration of two types of approach for materials design, the data-driven approach based on the comprehensive materials database and the model-driven approach based on the theoretical calculation corresponding to various physical models, has been proposed as an efficient way to accelerate the speed of finding target atomic configuration from huge number of candidates when designing new materials with specified properties. Mapping approach is utilized by mining the data in materials databases for discovering the regularities and correlations on the compound formation, constitution and crystal structures, which directly provides hints on candidate materials in preliminary stage. On the basis of regularities revealed by data mining, various theoretical approaches are applied to the target substances for investigating further their structural stability, phase equlibrium, and physical properties in order to estimate the possibility to be new materials and to understand the insight into the origin of those regularities. Several examples are presented to show how this combined approach is attempted towards discovery new materials.