It feels good to be back and posting on the UAH Honors SAGA blog. This post I will be focusing on one of my passions consuming a lot of my time in Oslo. On the academic side, I have been challenged with Senior level programming courses in Biologically Inspired Computing and Computational Physics. Coding has always been a passion for me and through my ME degree at UAH I have constantly felt the desire to learn advanced ML, AI, and randomization methods (Monte Carlo), and the University of Oslo gave me the chance. I have been able to keep my coding skills up by TA’ng ENG101 for three semesters (including the first Honors section!) but I felt it was time to further my skill set.
One aspect of programming Oslo that is like no other is the free wifi absolutely everywhere in public places and it’s super fast. This allows me to work on my projects with views like you can see below.
It is amazing what a relaxing environment does for productivity. I try and do this at UAH, but Oslo has definitely inspired me to come back and spend more time studying in places like Monte Sano.
Biologically Inspired Computing focuses on Neural Networks and Machine Learning techniques. There have been two mandatory assignments so far and they have definitely challenged my programming skills. This is mainly because I had to receive special permission to take them, due to the fact that they are Senior level Computer Science classes.
The first assignment involved using a genetic algorithm (GA) to solve the traveling salesman problem. By essentially breaking the problem into genotypes and phenotypes, you can use a genetic-based reproduction structure to develop your family bloodline. The GA then produces new generations of children to optimize the results towards the optimal solution.
The second project I just turned in last week was a Machine Learning classifier for EMG signals. The assignment explores supervised learning with the implementation of a multilayer perceptron (MLP). The task at hand was to analyze data from electromyographic (EMG) signals to learn eight hand gestures from a robotic prosthetic hand controller. The algorithm used backpropagation to fill the neural nodes to refine a weighting system for the dynamic classification process. The confusion matrixes are filled with the results allowing for analysis.
Both of these assignments allowed me to learn in-demand algorithm knowledge, as AI / ML is the future of all computing. You can view my work for the course on my Github course page and there are reports included in both for an easy to read summary. I feel blessed to have the opportunity to experience this world-class course that is recognized for it’s rigorous but exceptional content. The applications in the field of robotics (my future career path) are endless and provide a solid foundation for future growth in programming. This is all I have for this week and in my next post, I plan on talking about Computational Physics!