Projects

All of the code bases presented here can be perused on my GitHub page.

Software

These are some of the software research projects I haven been working on.

• Brownies.jl and dbfort are written to perform Brownian Dynamics simulations to study dynamical properties of hard-sphere systems. They work almost identically except that the former is written in Julia and the latter is written in Fortran.

• Cubed.jl attempts to be a CUDA implementation of the above project. This research code was meant as a learning project on how to employ Graphical Processing Units as means of faster computation.

• Newtman.jl is my side project on metaheuristics, a type of optimizers based on randomness and general rules to solve non-linear optimization problems. This project started with the purpose of using them to tune hyperparameters in my other Machine Learning projects.

• segmed, written in Python, is a research project focused on the ability to perform image segmentation using modern Deep Learning techniques. This project is quite complete as it contains robust unit testing and good software practices.

• Elysivm.jl, the newest addition, is an implementation of Least Squares Support Vector Machines. This type of Machine Learning models are a reformulation of classic Support Vector Machines, but instead of solving a quadratic convex optimization problem, they solve a least squares optimization problem, which is faster.