I went to a Machine Learning 101 event hosted at the Princeton Public Library. This meet up consisted of a long lecture on the basics of machine learning and then I introduced myself to attendees afterward and had discussions with them. In this lecture, I was taught basic concepts of machine learning including gathering data, abstracting this data, training a model on this data, and then improving the model as necessary. I also learned about types of machine learning algorithms including reinforcement, supervised, and unsupervised algorithms. The primary topic I wanted to discuss with others was about Magenta Studio, Google’s machine learning program for generating new music, which I am using in my thesis project. 

From talking about my thesis project with others at the meetup, I was briefly told about how Magenta works involving how it uses the MusicVAE model. It also was trained on millions of musical phrases and patterns that are primarily of Western origin. I was also told about more machine learning music creation tools like Amper Score, Melodrive, and Musico. When discussing machine learning software for music visualization, there were less well-known projects, but one I was told about is called APEXvj. Another topic I got to discuss a little bit was how machine learning can be used in web development. This short discussion ended up focusing on YouTube’s algorithms to detect content against their terms of service and how frameworks like Tensorflow can be used in conjunction with Python for web development.

This meetup impacted my thinking about my thesis project by introducing me to new machine learning technologies that I can possibly use for my thesis project. I also gained a new understanding of how machine learning works including how Magenta Studio functions so I can use it more accurately when creating new musical ideas. I was also surprised to learn of the various machine learning tools there are enlightening my perspective of what my thesis project can become with possibly learning frameworks like Tensorflow to enhance my thesis project and future career.