MACHINE LEARNING MODELS FOR THE PREDICTION OF TREATMENT CONDITION & POSSIBLE OUTCOMES

Objectives:
This WP addresses the development of Machine Learning models, which will firstly help to understand the relationships between changes in the physical-chemical and mechanical properties of the selected materials and the CO2 experimental conditions used. In addition, the use of computational modelling will enable to further tailor and optimise the CO2 experimental conditions for plastic materials, as well as foresee possible outcomes.
When successful predictions will be attained, the Machine Learning model will be published in the form of an Open Source software application, with the aim of assisting conservators with the selection of the optimal CO2 treatment conditions.

Tasks:
Task 3.1 - Population of a database with all the experimental data collected in WP2.

Task 3.2 - Implementation and training of Machine Learning algorithms.

Task 3.3 - Development of an Open Source software application for guiding conservators with the selection of the most suitable CO2-based treatments.

WP 2.

WP 4.