Artificial Intelligence reasearcher

My research interests lie in the Mathematical foundations that accounts for the booming success of Deep Learning techniques, and most generally AI. My recent work has taken a shift towards Computer Vision and Medical Imaging: Medical Image Segmentation and Registration.

In this blog, you will find some posts that I began writing during my master. Although it was mostly focused aroung documenting what I learnt, I am now thinking I would like this to be more open. If you have any ideas, shoot me an e-mail.

Note: Some links are still very much a work of progress. They are annotated with an asterisk *.

VisionNatural Language ProcessingNetworks science
Image-formation model ๐ŸŒ Language model ๐Ÿ“•Graph theory โ™ป๏ธ
Filtering ๐Ÿ“ทWords Representation โœ’๏ธ + codeRandom Network ๐ŸŒ
Canny edge detection โœ๏ธTagging and NERScale-free property โค๏ธ
Feature selection + stitch images ๐Ÿ’ ParsingBarabรกsi-Albert Model ๐ŸŒŒ
Optical flowMining User Generated ContentCommunity detection ๐Ÿ‘ช
Segmentation & GroupingMachine Translation and Natural Language Generation*Link prediction + Code ๐ŸŒฟ
Object Detection/ RecognitionMachine Readingย 
TrackingDialogueย 
* First order scattering transformโ€ฆย 
Machine LearningDeep LearningTools
Machine Learning basics ๐ŸŽฏMulti-Layer Perceptron ๐Ÿ’ปOptimization basics โœ…
Linear regression ๐Ÿ“ˆย Optimizing DL Models โญ๏ธย Unconstrained optimization ๐ŸŒ€
Logistic Regression ๐Ÿ“ŠBasics โœ…Constrained optimization ๐Ÿ“
Probabilistic classifiers ๐Ÿ”ฎ* Auto-Encoder โณEigenvalue decomposition ๐Ÿ’ 
LDA ๐Ÿ“* Word embeddings โœ’๏ธ + codeAnomaly detection โ—๏ธ
PCA basics ๐Ÿ”ŽBag-of-Words โœ’๏ธAnomaly detection time series ๐Ÿ•œโ—๏ธ
Naรฏve Bayes ๐Ÿ” CNN ๐ŸŒ„Feature engineering for music ๐ŸŽผ
* Decision trees ๐ŸŒฒ* Useful tricks and layers โžฟImaging time series ๐Ÿ’ˆ
Ensemble methods ๐ŸŒฒ๐ŸŒณRNN/LSTM ๐Ÿ•งMap Reduce โœ…
Support Vector Machine โ†—๏ธAvanced RNNs ๐Ÿ“˜Hadoop ๐Ÿšˆ
Clustering ๐Ÿ‘ซGAN & VAE ๐ŸŽญSpark ๐Ÿš… + notebook
โ€ฆย Statistics - Basics
โ€ฆย Statistics - Parameter estimation
โ€ฆย Statistics - Hypothesis testing
Artificial IntelligenceReinforcement LearningDecision Modelling
Problem solvingIntroductionโœ…Utility theory ๐Ÿ‘Œ
Adversial searchDynamic Programming ๐Ÿ’ปMCDA ๐Ÿ“‘
Constraint Satisfaction ProblemsPredictionsVoting and social choice theory ๐Ÿ—ณ๏ธ
โ€ฆControlSynthetis criterion ๐Ÿ†•
ย ย Outranking methods ๐Ÿ†

Highlight of my projects

  • Multi Modal Medical Image Registration
  • Organs at Risk Segmentation
  • Identify the genre of a song (from raw audio)
  • Infer social circles on social networks
  • Imaging time series
  • Improve predictive power for consumer behavior in the retail industry
  • Predict Estimated Time Arrival and delays of truck deliveries

Climate change

Outside on what I can learn about data science / AI related topics, I care a lot about climate change and the impact it will yield for earth and humans. Beside trying little things at my scale (biking to work, working on my vegetable garden, consuming less and betterโ€ฆ) I do the best to educate myself and others on what will probably be the biggest challenge the world need to face.

Get in touch for a collaboration!

If you are doing great work on a project needing Artificial Intelligence to move forward, shoot me an e-mail. I would gladly discuss it with you. From startups to pure research, I am interested in a wide range of opportunities.