- Prizes and Awards,
- Research,
Post-doc Tony Ribeiro receives the best paper award at the 30th International Conference on Inductive Logic Programming
Tony Ribeiro, a post-doctoral student in the LS2N Formal Methods for Bioinformatics research group - in the framework of a Franco-Japanese collaborative project co-financed by the National Institute of Informatics (NII, in Tokyo) and the RFI Atlanstic 2020 -, received the "best paper award" at the 30th International Conference on Inductive Logic Programming (ILP)
on November 15, 2021
The International Conference on Inductive Logic Programming) is a major international forum in the field of inductive logic programming, a sub-field of machine learning. This year, it was integrated into the IJCLR (International Joint Conference on Learning & Reasoning), which was held online from 25 to 27 October 2021.
Tony Ribeiro started this work in 2020, spending 6 months in Katsumi Inoue's team at the National Institute of Informatics (NII), and has recently had a paper accepted in the Machine Learning Journal. Since January 2021 he has continued his work as a post-doc in the MeForBio research group.
His contribution is in the field of learning dynamic models of biological systems. Until now, in order to build a discrete model from time series data (e.g. gene expression data as a function of time), it was necessary to make a prior hypothesis on the model's update scheme. This question of update schemes is at the heart of much research work, and among the best known are synchronous (several model variables are updated simultaneously) and asynchronous (a single variable can be updated) schemes.
In this paper Learning any memory-less discrete semantics for dynamical systems represented by logic programs (pre-print accessible via Hal), Tony and his co-authors proposed an algorithm for learning a logic model from time series data without making any prior assumptions about the update scheme of the model. This algorithm is accompanied by practical functionality, available to all via a Python API.
Acceptance of this paper in the Machine Learning Journal (ranked Q1 in Scimago) and as well as the "best paper award" highlight the paper's contribution.
Tony Ribeiro started this work in 2020, spending 6 months in Katsumi Inoue's team at the National Institute of Informatics (NII), and has recently had a paper accepted in the Machine Learning Journal. Since January 2021 he has continued his work as a post-doc in the MeForBio research group.
His contribution is in the field of learning dynamic models of biological systems. Until now, in order to build a discrete model from time series data (e.g. gene expression data as a function of time), it was necessary to make a prior hypothesis on the model's update scheme. This question of update schemes is at the heart of much research work, and among the best known are synchronous (several model variables are updated simultaneously) and asynchronous (a single variable can be updated) schemes.
In this paper Learning any memory-less discrete semantics for dynamical systems represented by logic programs (pre-print accessible via Hal), Tony and his co-authors proposed an algorithm for learning a logic model from time series data without making any prior assumptions about the update scheme of the model. This algorithm is accompanied by practical functionality, available to all via a Python API.
Poster session with the team at the ILP conference
Acceptance of this paper in the Machine Learning Journal (ranked Q1 in Scimago) and as well as the "best paper award" highlight the paper's contribution.
