Gerardo C.

Data Science Programmer

I am a Scientist researcher developer with experience in Neuroscience (PhD) and Biomedical Engineering (Msc.). My best skill and passion is programming data science solution in Matlab, R or Python. I have worked on Pattern recognition in time series, Brain-Computer interfaces and Computer Vision. I have programmed in Matlab, Python, R and C++ for testing many algorithm related to digital signal and image processing. The main projects I have worked on are: 1-Pattern recognition in time series, specifically in Capillary Electrophoresis data using Dynamic Time Warping and Wavelet analysis. The aim was developing of Matlab based software to achieve the alignment of many electropherograms with non-linear amplitude and time distortion in order to automate peak measurements. 2-EEG based Brain-Computer Interfaces. In my doctoral studies I did an internship at Dr. Guan Cuntai's Lab in the Institute for Infocomm Research in Singapore. I acquired experience in self-paced motor imagery detection using Common Spatial Patterns and SVM. Additionally, I worked on Visual Event Related Potential Classification to improve the P300 speller paradigm. 3-Feature extraction from Face videos and Physiological signals from Mahnob and DEAP dataset using SNN for emotion recognition. Face detection, Facial landmarks, HRV, frequency respiration, Temp, GSR, etc. 4-Time series pattern recognition and clustering in trading signals (candle sticks patterns) in R. Quantization, Documentation, Packaging, Vignettes, Rmarkdown. I taught Digital Biosignal and Image Processing Courses. I’m very familiar with several digital signal, image processing and pattern recognition techniques like wavelets analysis, sparse representation, sum of Gaussian representation, dynamic time warping, Artificial Neural Network and Support Vector Machines. Recently, I am very interested on Computer Vision. I have worked on automated measurement of the flexion of upper extremities using video analysis by contours tracking, I have also tried Deep Learning libraries (CNN) for object recognition using OPENCV in python.