Hi,
I am Rubbermaid Laverde, Master's candidate in Electronic Engineering with a focus in Bioengineering and Machine Learning.
I have 5 years of experience working in Machine Learning and all kind of electronic applications, including Arduino and Raspberry Pi programming. Over the last few years, I have developed projects such as: automatic oil-palm detection system based on a logistic regression model using UAVs, breast cancer detection using neural networks and automatic heartbeats detection. Currently, I am working on a project using neural networks for classification and automatic assessment of technical skills in laparoscopic surgery.
The knowledge that I have developed in my career over the last few years makes me an excellent candidate for your project. Do not hesitate to contact me and tell me more about the project details. I will be glad to work with you.
PD1: The training procedure of the ANN (or SVM) can not be performed directly in Arduino, at least not in a fast and simple way because the existing libraries of machine learning for Arduino are very limited. However, it is possible to train a more robust ANN using MATLAB and then deploy the trained ANN to the Arduino for its implementation.
PD2: If it is strictly necessary that the ANN training be performed directly in Arduino, I can do it using Neurona (Arduino Library), but this could take much longer and the ANN performance may not be the most optimal.