Dimitri Lozeve

Hi! I am a Research Engineer at Sysnav, where I work on developing algorithms to help physicians measure the progress of neuromuscular diseases. We use data from inertial sensors to extract meaningful variables for pharmaceutical studies.

I was previously a Data Scientist at Mindsay, working on natural language understanding for chatbots.

See my full CV, or a PDF version.

You can contact me by email (at firstname@lastname.com), reach me via Twitter, find me on LinkedIn, or see what I do on GitHub.

minisign/signify(1) public key: RWQ6uexORp8f7USHA7nX9lFfltaCA9x6aBV06MvgiGjUt6BVf6McyD26 (download)

Recent Posts

Get notified of new posts via RSS or Atom feeds!

• Online Analysis of Medical Time Series - November 17, 2020
• Learning some Lie theory for fun and profit - November 14, 2020
• Quaternions: what are they good for? - November 9, 2020
• Dyalog APL Problem Solving Competition 2020 — Phase II - August 2, 2020
• Dyalog APL Problem Solving Competition 2020 — Phase I - August 2, 2020
• Operations Research and Optimization: where to start? - May 27, 2020
• ICLR 2020 Notes: Speakers and Workshops - May 5, 2020
• Reading notes: Hierarchical Optimal Transport for Document Representation - April 5, 2020
• Mindsay: Towards Self-Learning Chatbots - April 6, 2019
• Random matrices from the Ginibre ensemble - March 20, 2019

Find more in the archives.

Some Projects

• Topological Data Analysis of time-dependent networks (sig)
  Application of the recent advances in Topological Data Analysis (TDA) and Persistent Homology to periodicity detection in temporal networks.
• Community Detection
  Community detection algorithms and applications to a large social network obtained from Prestashop data.
• Ising model simulation
  Monte Carlo simulation of the Ising model using Metropolis-Hastings.
• GardenOptim
  Optimal allocation of plant species on a non-rectangular garden subdivided in individual cells, with constraints from an affinity matrix between neighboring species. Uses mixed integer quadratic programming and Markov chain Monte Carlo.

More on the Projects page.