For registration, please fill the necessary data in the registration section of the event website: https://ml4ms.ijs.si/registration/

Deadline for abstract submission: 29.02.2024
Notification of acceptance: 13.03.202

When the exceptional properties of a newly engineered material are discussed in the literature it is common to present:
- a text-based description of the sequence of actions through which such material was obtained, listing key variables as scalars. 
- a characterization of its structure by means of advanced microscopy (e.g., 2D images, 3D tomographies, 4D spatio-temporal analysis) and spectroscopy (e.g., adsorption spectra, NMR spectra), also with the aid of atomistic and electronic structure simulations. 
- a list of key performance indicators, in the form of scalar variables (e.g. the mechanical properties of an alloy or the Seebeck coefficient of a thermoelectric) or a time-series (e.g., activity of a catalyst over time, the capacity of a batter over time).
- a mechanistic discussion of the relationships that link structure-to-property, often through quantities extracted from electronic structure and atomistic scale simulations.
A rooted knowledge and understanding of a material and its properties thus stems from a holistic perspective.

In this context, machine learning methods have emerged as revolutionary tools to accelerate materials design and discovery. They found a use in each of the specific tasks discussed above, and notable examples indeed include (but are not limited to):
- Text-mining coupled to regression and optimization for materials synthesis [1,2].
- Advanced method to match materials structures/phases and spectra [3,4,5].
- Automated detection methods for images [6].
- Surrogate models for property prediction and high-throughput screenings [7,8].
- Symbolic regression and interpretable approaches for descriptor identification [9,10].

Through the application of these methods, stand-out successes have been achieved in the data-driven discovery of, e.g., better catalysts for small molecules (electro)chemical conversion [11,12], high-performance energy-storage [13], and energy-conversion [14] materials. Similarly, fundamental bottlenecks in the modelling and simulation of complex interfaces,[15] in the characterization of advanced materials,[16] in the synthesis of high-value compounds [17] have been overcome.

Linked to the continuously evolving development of machine learning methods for materials science, challenges in capacity-building and research-coordination naturally arose. Two main objective of this school(first four half-days)+workshop(last four half-days) are then identified:
- Young researchers will have the opportunity to grow solid foundations and a complete overview of the cutting-edge approaches that enable the community to tackle outstanding challenges across diverse domains in materials design and discovery. To this end, during the school part, renowned experts will discuss state-of-the-art machine learning methods and applications across the full breadth of the materials design-make-test-analyze cycle, under a pedagogical tenet.
- Attendees will have the chance to discuss and identify routes on how to best combine information of different nature such as data from simulations and from experiments, images and text or scalar variables, 2D images or 1D spectra to materials properties, towards a unified vision (and solution) of the material design and discovery problem. To this end, during the workshop part, invited and contributed speakers, and panel discussions will take place, with a focus on multi-modal, multi-objective, and multi-fidelity machine learning methods in materials science.