Christian Jauvin

Independent software developer / data scientist / researcher (M.Sc. CS + Machine Learning, 2003, UdeM).

Ongoing

My current business model is to partner up with researchers in various fields to help them solve problems with computing techniques (machine learning algorithms, and other highly transferable skills), open source software and tools, and sometimes even come up with original research ideas and directions. Currently I’m working with:

• Direction de santé publique de Montréal
  • Road accident address correction and geocoding (with Dr. Patrick Morency)

Non-scientific

• Société Roucet Inc.
  • Web-based ERP application (built with Python and Ext JS) for a wine importing company

Past

• McGill Surveillance Lab
  • PopHR: a semantic web application for the computation, management, querying and visualization of public health indicators (with Dr. David Buckeridge)
  • Gastrointestinal outbreak simulation on a CLUMEQ supercomputer
  • Rotavirus vaccination data management system
  • NLP-based social media monitoring of vaccination refusal
  • Machine learning experiments for the analysis and classification of CIHR grant applications (with Dr. Robyn Tamblyn)
• Institut québécois d’aménagement de la forêt feuillue
  • Botanical tree reconstruction from LIDAR point cloud data (PypeTree)
• Real-time spatial surveillance application (Direction de santé publique de Montréal)
• Montreal Chest Institute
  • Tuberculosis data management system
• Residential address error correction algorithm used for a study of their effect on public health data geocoding
• Deep vein thrombosis (DVT) textual diagnostic machine learning classifier
• PostGIS training
• Generative art with the Capture collective

Code

• PypeTree, a Python/VTK-based graphical tool to reconstruct botanical trees from point cloud data (we published an article about it)
• Dracones, a web mapping framework based on MapServer
• Various smaller projects and experiments (@GitHub)

Publications

• Delagrange S, Jauvin C, Rochon P. PypeTree: A Tool for Reconstructing Tree Perennial Tissues from Point Clouds. Sensors. 2014; 14(3):4271-4289.
• Shaban-Nejad A, Jauvin C, Lavigne M, Izadi M, Mondor L, Okhmatovskaia A, Buckeridge D (2013). PopHR: An Integrated Semantic Framework for Population Health Surveillance. To appear in Proc. of The 4th International Conference on Biomedical Ontologies (ICBO 2013), Montreal, Canada, 8-9 July 2013.
• Shaban-Nejad A, Okhmatovskaia A, Izadi M, Naderi N, Mondor L, Jauvin C, Buckeridge DL (2013). PHIO: A Knowledge base for Interpretation and Calculation of Public Health Indicators. to appear in Proc. of the 14th World Congress on Medical and Health Informatics (MedInfo 2013),Copenhagen, Denmark, 20-23 Aug 2013.
• An infrastructure for real-time population health assessment and monitoring, Buckeridge D, Izadi M, Shaban-Nejad A, Mondor L, Jauvin C, Dube L, Jang Y, Tamblyn R, IBM Journal of Research and Development, 2012.
• Simulation Analysis Platform (SnAP): a Tool for Evaluation of Public Health Surveillance and Disease Control Strategies, Buckeridge DL, Jauvin C, Okhmatovskaia A, Verma AD, AMIA Annual Symposium, 2011.
• Residential address errors in public health surveillance data: A description and analysis of the impact on geocoding, Zinszer K, Jauvin C, Verma A, Bedard L, Allard R, Schwartzman K, de Montigny L, Charland K, Buckeridge DL, Spatial and Spatio-temporal Epidemiology. 1(2-3): 163-168, 2010.
• Quantifying the Potential Benefit of Early Detection for Preventing Morbidity and Mortality: A Simulation Study of Cryptosporidium Outbreak. Okhmatovskaia A, Verma AD, Jauvin C, Barbeau B, Allard R, Buckeridge DL, presented at the ISDS Conference, 2010.
• A Neural Probabilistic Language Model, Yoshua Bengio, Réjean Ducharme, Pascal Vincent and Christian Jauvin, Journal of Machine Learning Research, 3:1137-1155, 2003.

Other

• My technical blog: Autotelic Computing
• My Twitter feed
• MOOCs (self-education)
  • I have completed the Fall 2011 Stanford online AI Class
  • I have completed (with a perfect score) the Spring 2012 Stanford online Cryptography Class
  • I have completed (with distinction, for a final score of 95.4%) the Neural Networks for Machine Learning 2012 Coursera class, taught by Geoffrey Hinton
  • I have completed (with distinction and a perfect score, which was quite hard to do!) the Discrete Optimization 2013 Coursera class, taught by Pr. Pascal Van Hentenryck.
  • I have completed the Computer Networks 2014 Coursera class, taught by Pr. David Wetherall.
• My email: cjauvin@gmail.com