Complex data analysis and applied artificial intelligence

Recent developments in the scientific topics addressed by the unit have been accompanied by a profound transformation in the nature, volume, and complexity of the data produced. Research work is now based on multi-criteria evaluation approaches that integrate heterogeneous data sets (experimental, environmental, phenotypic, biological, and molecular data), often on a large scale and generated by increasingly instrumented experimental devices.
The analysis and exploitation of these data now require the implementation of advanced methods of statistical analysis, scientific computing, and machine learning, as well as the development of reproducible and robust processing chains. Traditional approaches are reaching their limits in the face of multidimensional data, high variability, and the need to cross-reference multiple levels of information, particularly in the case of biological and molecular data.
In this context, the use of innovative data analysis methods, including approaches based on artificial intelligence, is becoming essential in order to extract relevant information, improve the robustness of results, and enhance the scientific quality of the unit's output. 

Several of the unit's projects use these innovative approaches. The SAFECLIM project aims to study the impact of climate change on food safety, focusing on the relationship between climate variations (high temperatures, droughts, extreme rainfall) and increased risks associated with Salmonella in France and Poland.

Theses supervised on this topic:

Maëva Caillat's thesis in the HEMIC project aims to enhance the Sym'Previus software through the semi-automatic collection and analysis of data from the literature (data mining, modeling, AI).

Contact persons:

• Rodney Feliciano, researcher at SECALIM
• Jeanne-Marie Membré, researcher at SECALIM