Climate change and risks 3 min

More efficient calculation models to serve agricultural research

Scientists at the INRAE Occitanie-Toulouse and Versailles-Grignon Research Centres have developed ELSA, a new method to extract knowledge from numerous observed data points. This method has already proved its worth by supporting geneticists in their search for future sunflower varieties that are more drought-resistant.

Published on 07 April 2022

illustration More efficient calculation models to serve agricultural research

Bayesian networks are simple, probabilistic graphical models, which were introduced in the 1980s and have been used increasingly widely since then. They are commonly used for classification and in the computer science fields of machine learning and artificial intelligence. They are both knowledge representation models, acting as “calculating machines” for conditional probabilities and decision support systems. The automatic learning of a Bayesian network from observations of random variables enables the extraction of knowledge from data.

Until the 2000s, the methods used were limited to around thirty variables. Major advances have been achieved during the past ten years thanks to integer linear programming and optimum cut generation, thus pushing the limit to around one hundred variables.

More rapid and more efficient

Scientists in the Mathematics and Applied Informatics Unit in Toulouse (MIAT) at the INRAE Occitanie-Toulouse Research Centre, and the Mathematics and Applied Informatics Unit (MIA) at the Versailles-Grignon Research Centre have developed a new and more efficient method which requires a shorter calculation time. Called ELSA, for Exact Learning of Bayesian network Structure using Acyclicity reasoning, this method picks up on the idea of cuts by integrating them in CPBayes, a constraint programming tool dedicated to this learning problem.

“Using a different approach, we have developed a high quality, and notably much more rapid, calculation method” explains Simon de Givry, research scientist in the MIAT Unit. “Using a test set of forty problems containing more than 500 random variables, and over the same allotted time of 90 hours, ELSA managed to optimally solve 23, while GOBNILP only solved nine and CPBayes just four.”

Among the prospects for further research improvements, the use of decision tree diagrams is expected to enable more efficient manipulation of very large value domains, which would further accelerate the operations performed by ELSA.

One of the tools used to design future sunflower varieties

In the context of the Sunrise programme, studies were carried out by the Joint Research Unit for Plant-Microbe-Environment Interactions (LIPME) to identify genes of interest regarding drought tolerance and to model the agricultural characteristics of future sunflower varieties carrying these genes.  ELSA contributed to better understanding the genetic and molecular bases controlling plant physiology and development in order to predict the characteristics of the hybrids.



Fulya Trösser, Simon de Givry, and George Katsirelos
Improved acyclicity reasoning for bayesian network structure learning with constraint programming
Proc. of IJCAI-21, Montreal, Canada, 2021

Learn more


The International Consortium on Sunflower Genomics

The International Consortium on Sunflower Genomics (ICSG), coordinated by INRAE Occitanie-Toulouse, joins 4 public labs from France, Canada, United States and Israël, and 8 private partners. During 4 years, the ICSG will characterize genetic diversity in sunflower by producing new high quality genome sequences of wild sunflowers. These data will improve our knowledge on the structure of the genomes and on their evolution and facilitate the use of wild relatives in breeding improved sunflower varieties.

10 January 2020

Climate change and risks

Using environmental scenarios to determine the world’s future

AllEnvi, the French national alliance for environmental research, has produced a study on possible future environmental scenarios based on a survey of major international foresight studies. The study tackled two issues: What do the scenarios look like and how long are their timelines? While future possibilities go from one extreme to another – from “Chaos” in the trajectories to “Positive Synergies” or environmental degradation and improvement – soil and water appear to face the most danger. These scenarios revealed major deficiencies with regard to seas, oceans, coastlines and forests that need to be quickly addressed.

23 January 2020

Food, Global Health

Early calculation of actual COVID-19 fatality rate in France made possible by mechanistic-statistical model

PRESS RELEASE - By the end of March, an INRAE team had already produced the first infection fatality ratio calculation for COVID-19 outside China. They identified a fatality rate of 0.5% based on French hospital data, adjusted to 0.8% when nursing home data were included. Their results were confirmed at the end of April by work carried out at the Pasteur Institute in Paris and by a New York study. The team’s analysis was published on 8 May 2020 in MDPI Biology.

12 May 2020