MOTIVATIONS & PURPOSES OF THE SCHOOL
Undergraduate and graduate students in Physics or Chemistry have, in general, a solid background in Statistical Mechanics and Condensed Matter Physics but, typically, very little (if any) knowledge of the numerical tools that are the bedrock for any theoretical work in such areas.
The proposed activity aims to fill this training gap by providing a comprehensive introduction to stochastic simulations techniques, without any prior knowledge of scientific programming. This makes this activity unique, different in particular from "traditional" schools directed to a more specialized, already trained, audience.
The purpose of this School is threefold:
(i) providing students with a basic but detailed overview of the theoretical foundations of stochastic simulation techniques, with particular emphasis on classical and quantum matter through Monte Carlo methods;
(ii) giving an overview of the domains of interesting applications;
(iii) providing the basics for writing and running in practice simple simulations using Monte Carlo codes.
At the end of the School, the students will:
(i) have acquired a first-hand experience with numerical simulations for a broad range of common scientific contexts;
(ii) be aware of the challenges and open problems in Computational Physics;
(iii) be able to write their own codes for simulating and testing systems of their interest.
PROGRAM of the SCHOOL (26 June - 13 July): Location & details
IMPORTANT: A pre-School (Tuesday, 26 June - Friday, 29 June), intended as a pedagogical introduction to numerical methods, will be offered during WEEK 0 preceding the 2-week Summer School. Notice that this pre-School is not mandatory, yet it is highly recommended to those who are not familiar with simple concepts of numerical computation. Instead, attendance for the whole period WEEK 1-2 is strictly compulsory. We will not accept candidates who wish to attend only one of the two weeks.
Each day, lessons are structured in two parts:
- 09:00 - 12:00 (Theory): Room #005 (Sissa ground floor)
- 14:30 - 17:30 (Practicum): Room #003 (Sissa ground floor)
WEEK 0 (pre-School, 26-29 June): Introduction to Numerical Methods
- June, 26: Introduction to Fortran90 - Part 1
- June, 27: Introduction to Fortran90 - Part 2
- June, 28: Modeling of Data
- June, 29: Integration of Ordinary Differential Equations
WEEK 1 (2-6 July): Monte Carlo Methods for Classical Systems
- Introduction to Probability Theory
- Direct Sampling
- Markov Chains
- Monte Carlo as an Optimization Tool: Simulated Annealing
- Advanced Sampling Techniques: Reweighting and Parallel Tempering
- Stochastic Processes: Wiener Processes and Brownian Motion
- Introduction to Molecular Dynamics: Integrators, Constraints, Thermostats and Barostats
- Hybrid Monte Carlo
WEEK 2 (9-13 July): Monte Carlo Methods for Quantum Systems
- Variational and Diffusion Monte Carlo for Atoms and Molecules
- Energy Derivatives and Stochastic Optimization Methods
- Calculation of Atomic Forces by Quantum Monte Carlo: Structural Optimization and Molecular Dynamics
LOCATION OF THE SCHOOL
The School will be held at SISSA (www.sissa.it), in Trieste (Italy). More details concerning the logistics (rooms, equipment, etc.) will be sent to successful applicants before the start of the School.
DEADLINE for APPLICATION
The closing date for applications is 8 April 2018.
IMPORTANT: Applications must be accompanied by the CV of the applicant.
- A registration fee (200 Euro) will be requested to the accepted candidates and reimbursed to students who will have fully attended the school.
- The organizers will cover accommodation (14 nights), lunches and dinners at SISSA (Monday to Friday) and public transportation. Lodging expenses will be covered also during the pre-school.