Designing new machine learning tools for high-dimensional time series and longitudinal data.
Applying deep reinforcement learning and computer vision in biomechanics.
A co-founder of DeepArt - the neural style transfer company.
Constantly looking for new opportunities and challenges 🚀

Download my latest CV

Recent projects


Statistical software for sparse longitudinal data

HealthAI working group

Bringing together engineering, medicine and business at Stanford (400+ members).

Deep learning in gait analysis

Modern computer vision in biomechanics

DeepArt - Neural style platform

Computational methods for transfering a painting style to any photo

Learning how to run (NIPS)

Neuromuscular simulations for studying motor control. Official NIPS challenge in 2017 & 2018.


Popular press coverage of my projects

TechCrunch IEEE GlobalSpec EPFL news Stanford news China daily Live Science Le temps 24heures Technologist Stanford SCOPE

Selected Publications

In clinical practice and biomedical research, measurements are often collected sparsely and irregularly in time while the data acquisition is expensive and inconvenient. We propose an alternative elementary framework for analyzing longitudinal data, relying on matrix completion.
Under review, 2018.

The academic and behavioral progress of children is associated with the timely development of reading and writing skills. Our method achieved 96.6% sensibility and 99.2% specificity. Given the intra-rater and inter-rater levels of agreement in the BHK test.
In npj Digital medicine, 2018.

Single-event multilevel surgery (SEMLS) is a standard treatment approach aimed at improving gait for patients with cerebral palsy. We identified 26% of the non-surgically treated limbs that may have shown a clinically meaningful improvement in gait had they received surgery.
In Scientific Reports, 2018.

Within the framework of functional data analysis, we develop principal component analysis for periodically correlated time series of functions.
Journal of Time Series Analysis, 2017.

In this paper, we address the problem of dimension reduction for time series of functional data ($X_t$: $t$ ∈ $Z$). Such functional time series frequently arise, e.g., when a continuous-time process is segmented into some smaller natural units, such as days.
In JRSSB, 2015.

Recent Publications

More Publications

. Longitudinal data analysis using matrix completion. Under review, 2018.

Preprint PDF Code Project

. Automated human-level diagnosis of dysgraphia using a consumer tablet. In npj Digital medicine, 2018.


. Estimating the effect size of surgery to improve walking in children with cerebral palsy from retrospective observational clinical data. In Scientific Reports, 2018.


. Principal component analysis of periodically correlated functional time series. Journal of Time Series Analysis, 2017.

Preprint PDF Code Project

. Orchestration Load Indicators and Patterns: In-the-wild Studies Using Mobile Eye-tracking. IEEE Transactions on Learning Technologies, 2017.

. Estimation in functional lagged regression. Journal of time series analysis, 2015.


. Dynamic Functional Principal Component. In JRSSB, 2015.

Preprint PDF Code Project

. How Do In-video Interactions Reflect Perceived Video Difficulty?. EMOOCs 2015, 2015.


. MOOC video interaction patterns: What do they tell us?. Design for teaching and learning in a networked world, 2015.


. Translating head motion into attention-towards processing of student’s body-language. EDM, 2015.



I was a teaching instructor for the following courses:

  • CS-411: Digital education & learning analytics, fall 2014, EPFL
  • MATHF309: Analyse Multivariée, spring 2014, ULB (in french)
  • STATF407: Stochastic Models, fall 2013, ULB
  • STATF407: Stochastic Models, fall 2012, ULB