AI4Science loader

Project Information

Project Overview

AI4Science project is a multidisciplinary initiative aimed at advancing the role of artificial intelligence in the scientific process. It focuses on developing novel AI methodologies—spanning explainable machine learning, foundation models, automated scientific modeling, and semantic technologies—to address the unique challenges of applying AI in the physical and life sciences. The project emphasizes the integration of data and domain knowledge, transparency in AI models, and support for open science principles. Applications range from drug and gene therapy design to equation discovery, environmental modeling, and materials science. With a strong consortium of Slovenian research institutions and access to state-of-the-art computational infrastructure, AI4Science seeks to significantly enhance scientific discovery through AI-driven tools and frameworks.

WP1: Explainable Machine Learning for Science
EExplainable Machine Learning for Science
WP2: Foundation Models for Science
Foundation models for Science
WP3: Automated Scientific Modelling
Automated Scientific Modelling
WP4: Semantic Technologies for Open Science
Semantic Technologies for Open Science
WP5: Project management, dissemination, and communication
Semantic Technologies for Open Science

Objectives

  • 1a. To develop explainable ML methods for interpretable modeling of complex data, integrating neural and symbolic approaches, explaining predictions, and tracking scientific trends using bibliographic data.
  • 1b. To apply explainable ML in complex settings to scientific problems, including trend monitoring, gene therapy, drug design, and mathematical discovery.
  • 2a. To develop methodology for pre-training and fine-tuning multimodal foundation models, to be used in different domains of science.
  • 2b. To learn and apply multimodal foundation models in different scientific domains, including medicine and healthcare, as well as materials science.
  • 3a. To develop AI methods to learn equation-based scientific models from data and domain knowledge using symbolic and neural approaches for accuracy and interpretability.
  • 3b. To apply equation-based AI methods to different domains: plant biology and ecology, as well as electrochemistry and materials science.
  • 4a. To develop semantic resources that describe ML and optimization and support AI applications across scientific domains.
  • 4b. To apply explainable ML to relate tasks/problems and properties/configurations of algorithms to algorithm performance in the areas of ML and optimization.
  • 5. To ensure effective project execution through strong coordination, quality management, open science practices, and broad dissemination to key stakeholders.

Timeline

Start: October 1st 2024
End: September 30th 2027.

Each Work Package (WP) includes two parts: developing AI methods and applying them in scientific domains, each divided into three tasks. Method development tasks usually start earlier and follow a cycle of design, implementation, and evaluation of AI methods. Application tasks begin with problem definition and data collection, followed by applying AI methods (often machine learning) to generate insights, and ending with evaluation from the perspective of the scientific field.