IPOL Journal · Image Processing On Line
Nicolas Limare will represent IPOL in the R4 meeting on reproducible research ("Rencontre de Réflexion autour de la Recherche Reproductible"), Thursday April 5th, at the Orleans University. Participation is free but registration is required.
Excerpt from the program:
Reproducibility in computational science now seems to capture everyone's attention. Considerations about the reliability of the numerical results published in scientific journals appear in diverse fields.
These efforts are essential to go beyond a credibility crisis in all disciplines where numerical results are important in the results published. No one today can guarantee the validity of most results presented at conferences and in research articles. Even the mature branches of science, despite their best efforts, sometimes suffer from problems of numerical errors affecting the final conclusions of the scientific literature. However, the traditional scientific publication alone does not determine the origin of these errors. This is why new standards of verifiability must be developed.
The objective of this first meeting is to compare points of view of scientific disciplines on the notion of reproducibility and identify various initiatives. Beyond the specific diagnosis for each discipline, it will identify some common issues and discuss solutions that can emerge around the concept of executable paper.
