Data Scientist at RocheDC 2020-Now
Implementation of predictive analytics algorithms as a tool for prognosis of diabetes related complications. Tools used; Python: Keras, Tensorflow, sklearn, Numpy, scipy, Pandas, Flask. AWS. Git.
Hi!
My primary interest lies at the cross-talk between computational neuroscience, artificial neural networks and machine learning.
I am currently working as a Data Scientist at Roche Diabetes Care, where I design predictive analytics algorithms to anticipate the onset of diabetes related complications, such as chronic kidney disease, heart failure or diabetic rhetinopathy.
Before joining Roche I worked in the Artificial Intelligence lab in the research unit from Telefónica Innovation, called Alpha. There I developed algorithms for machine learning interpretability, with applications to image recognition, natural language processing or data mining.
Prior to joining Telefónica, I was a member of the Neurophysics lab of Prof. Jordi Soriano in the physics department of the University of Barcelona. We conducted data analysis and modeling of neurodegenerative diseases, specifically the analysis of calcium fluorescence traces of induced Pluripotent Cells (iPSCs) derived from patients suffering from Parkinson’s disease, and the simulation of the dynamics of pathologic neuronal cultures taking into account experimental evidence.
Strong math background thanks to my undergraduate and graduate studies in physics
Strong physics background acquired during my undergraduate and graduate studies in physics
Programming dexterity and knowledge of diverse programming languagues, ranging from Python to C/C++ through Javascript and Fortran.
Background in complex systems thanks to the work conducted during PhD and early post-doc.
Advanced knowledge of neuroscience topics gained during my PhD studies
Currently involved in software development at my company for product delivery
Here you can find a brief self-assessment of skills quality and strength. Please note this has not been scientifically proven!
Programing skills
Soft skills
Degree obtained from the Universitat Politècnica de Catalunya.BarcelonaTech. Specialised in computational neuroscience, neural networks, network science and complex systems. Studied the computational properties of collective neuronal oscillations emerging from large scale neuronal networks at different scales.
Used several computational techniques and programming languages. Conducted large scale simulations using Fortran and C/C++, and the analysis of large amounts of resulting data.
Degree obtained from the Valencia International University. Specialised in secondary school teaching (physics and chemistry). Practical sessions in a secondary school, teaching physics, maths and chemistry to students with ages comprised between 12 to 18 years old.
Degree obtained from the Universitat de Barcelona. Specialised in computational neuroscience and computer simulations. Conducted experiments in the laboratory of Prof. Jordi Soriano on neuronal cultures grown from rat embryos. Studied the effects of electrical stimulation on the patterns of activation of the cultured neuronal networks.
Degree obtained from the Universitat Autònoma de Barcelona. Specialised in theoretical physics (quantum physics and biophysics). Bachelor's thesis on the role of astrocytes and myelin on the signal transmission properties along the axons of cortical neurons.
Implementation of predictive analytics algorithms as a tool for prognosis of diabetes related complications. Tools used; Python: Keras, Tensorflow, sklearn, Numpy, scipy, Pandas, Flask. AWS. Git.
Research on machine learning interpretability applied to computer vision, NLP and tabular data. Development of interpretable recommender sys- tems based on matrix factorization. Data analysis and visualization through the creation of interactive dashboards. Tools used; Python: Pytorch, Tensorflow, sklearn, Numpy, scipy, Pandas, Flask. AWS. Spark. Git.
Data analysis of the activity of induced Pluripotent Stem Cells (iPSC) de- rived from Parkinson’s disease affected patients. Large scale simulations of neuronal networks. Tools used; Matlab, Python (sklearn, Numpy, scipy, Pandas), UNIX (Bash), C/C++.
Agent-based modelling of Drosophila larvae chemotaxis. Data analysis or larval chemotaxis. Simulations of larval behaviour. Robotics (kilobots) and swarm behaviour (in collab. w. Dr. James Sharpe’s lab at CRG). Tools used; Matlab, Python (Numpy, scipy).
Organization of the physics laboratory lectures for undergraduate students in Industrial and Aerospace engineering. Problem solving instruction. Exam surveillance and correction.
In this short note I try to explain what Machine Learning is and what are the most important types of it.