Development of an algorithm to simulate functional images.
Software Developer, Matlab, PCA/ICA, KDE · In this work, we propose the simulation of new functional brain imaging samples as a solution to the small… · More sample size problem. The system takes existing functional imaging datasets, and extracts commons features using Principal Component Analylis (PCA). Afterwards, the statistical distribu- tion of the coordinates in the eingenbrain space is estimated usign Kernel Density Stimation (KDE) and new coordinates that match that distribution can be used to generate new images with similar properties to the original ones. Our strategy is intended to serve as a computational aid to reduce bias in the evaluation of new Computer Aided Diagnosis systems.
I coded and documented all the algorithms used, from a given real dataset of brain images.
What I Do
Data scientist at OrbitalAds, a startup focused on the automation of online marketing, with the aim of generating perfect ads through the use of Machine Learning and Natural Language Processing.
My role at OrbitalAds is designing and develop new ML algorithms with NLP.