Many of the projects presented here have been edited or taken out of their most practical context for this overview.
Python
: pandas, numpy, lightgbm, matplotlib
Entire workflow of the construction of a content-based recommendation model for movies using Gradient-Boosting Machine. From acquiring the data, preprocessing it, visualizing the statistics, constructing the feature for the model and finally constructing a recommendation model.
Python
: FEniCS, SciPy, Matplotlibgmsh
This work integrates meshes generated with the gmsh tool with the FEniCS library, used to solve PDEs with Python through finite elements. At first, the dynamics of a vortex dipole is simulated by means of a linearized Navier-Stokes direct resolution (baseflow.py). Then, the tools learned in class for spectral analysis of flow stability and optimal perturbations are used (EVP.py).
Python
: SciPy, Matplotlib
Files for research projet on flow instability develloped at PMMH with Laurette Tuckerman
- Heat equation solved with finite difference method in space and runge kutta 4.5 in time
- Couette flow equations proposed by Dwight Barkley "Modeling the transition to turbulence in shear flows" implemented in finite difference with runge kutta 4.5 in time
Python
: cv2, scikit-learn, pandas, Matplotlib
Basic computer vision models and operations like: convolutions, feature detections, Harris point of interest, k-means clustering, feature matching, PCA
Python
: odeint, Matplotlib animation, 3d, numpy
Solution of Lorenz's system of ordinary differential equations and visualization
Python
: scipy, scikit-learn, pandas, seaborn, pandas, matplotlib
Restricted Boltzmann machine (RBM) algorithm done for the course IA304-Probabilistic Models and Machine Learning at Telecom Paris
Many other projects were put aside for confidentiality reasons (as in the case of the internship I did at ONERA Aerospace Laboratory or ISQ), or because of my carelessness in having lost them.