Introduction
The LASSO group is an academic research laboratory specializing in
- Learning Algorithms: fundamental new algorithms that learn how to make prediction from large data sets.
- Statistical Software: new free/open-source software packages for large data analysis.
- Optimization: algorithms based on ideas from convex optimization (gradient descent) and discrete optimization (dynamic programming).
Machine learning is a branch of Artificial Intelligence which is concerned with algorithms for large data. Our lab works in this area, with particular expertise in algorithms for analysis of scientific data, from fields such as genomics, neuroscience, medicine, microbiome, cybersecurity, robotics, satellite/sonar imagery, climate/carbon modeling.
The acronym “LASSO” is a reference to the highly influential paper by Tibshirani, “Regression shrinkage and selection via the lasso,” JRSSB (1996), which is an inspiration for our research into interpretable machine learning.
Selected funded projects
- “Efficient algorithms and software for change-point detection.” PI Toby Hocking, travel grant from DATAIA (Artificial Intelligence Institute of Université Paris-Saclay), 25,000 euros, Academic year 2024-2025.
- “POSE: Phase II: Expanding the data.table ecosystem for efficient big data manipulation in R.” PI Toby Hocking, National Science Foundation grant 2303612, from program Pathways to Enable Open-Source Ecosystems (POSE), US$731,881, Sep 2023-Aug 2025. Co-PIs Igor Steinmacher and Marco Gerosa.
- “Addressing Structural Disparities in Autism Spectrum Disorder through Analysis of Secondary Data (ASD3).” PI Olivia Lindly, Co-PI Toby Hocking, National Institute of Mental Health grant 3R01MH134177-02S1, US$455,660, Aug 2023 to June 2027.
- “MIM: Discovering in reverse – using isotopic translation of omics to reveal ecological interactions in microbiomes.”, PI Jane Marks, Co-PI Toby Hocking, National Science Foundation grant 2125088, from program URoL-Understanding the Rules of Life, US$3,000,000, Sept 2021 to Aug 2026.
- “RcppDeepState: an easy way to fuzz test compiled code in R packages.” PI Toby Hocking, R Consortium Grant, US$34,000, Jan-Dec 2020.
Who we are
Toby Dylan Hocking
Professeur Agrégé / Tenured Associate Professor, Université de Sherbrooke, Département d’informatique.
You?
If you are interested to join the LASSO Lab, please read the application instructions.