Looking Back at Semester 1

       My group is under Dr. Hassibi for Machine Learning. This semester we have studying Machine Learning but more specifically: graph clustering algorithms. We tested a lot of algorithms on Zachary's Karate Dataset but what was more important than implementing them was understanding them. A lot of the algorithms were math intensive so in order to really understand why algorithms work the way they do, we focused on explaining the math. In order to see how many options there are to cluster the Karate dataset, algorithms were assigned after each meeting. Some algorithms that I did were Spectral Clustering, Girvan Newman, and more recently Fuzzy c-means.

Karate Club Dataset

       Aside from learning new algorithms, I think my Caltech experiences have helped me improve in other ways. For instance, when I began, I did not know a lot of linear algebra, but after studying math-centered subjects for a couple months, I have gotten into the habit of teaching myself it. I also know a couple of ways to professionally report results/teachings such as a detailed research report and a technical journal. Along with new methods of learning like creating concept maps, we made videos of ourselves explaining clustering algorithms for the public to view which was definitely out of my comfort zone but nevertheless super rewarding. But most importantly, I learned that I was passionate about Artificial Intelligence throughout this semester. I spent my summer not really enjoying ML but this Caltech research class has pushed me to do more outside of school.

Our research group

       This semester I really enjoyed the lectures from Dr. Hassibi. While he only showed up for 2 of the 5 meetings, they were insightful. However, much improvement is needed because the lectures were not structured nor were they particularly helpful for me to do research. The meetings are the most important part of this class because we get to meet a professional. To see that so much time was poorly spent on us summarizing what we learned for an hour was painful. I definitely believe that much can be done to improve this. This brings up another aspect of this class that I liked: the research was individual and we had to search for our answers. Dependence on teachers is an apparent issue because students become less self-reliant and thus fail on learning more without the help of others. While letting us go out to learn without training wheels seems like a lazy way to manage a class, in this class, it was much needed and I appreciate it. Additionally, I enjoyed working in a group because it created a friendly learning environment. I didn't enjoy concept maps or technical journals because a lot of what we wrote was repetitive and unnecessary. Another aspect of this class that came to a surprise to me was the lack of math we needed to learn. During my summer program for Artificial Intelligence, we were expected to know linear algebra and so spent some time to learn the basics. This class just jumped into the algorithms and so much of the time spent understanding them was wasted relearning math that should have been gone over at the start.

Fuzzy C-means is what I studied

       This class was definitely not what I expected. As mentioned before, we headed into research without actually understanding linear algebra. I also want to re-highlight the fact that the meetings with Dr. Hassibi, the honest reason why we wanted to take this class, were downright useless. We should spend valuable time letting him talk, not us summarizing what we learned. I get that him correcting our understanding is essential, but to me, I think there exists a more beneficial and efficient way. I also expected a better connection between the students and the professor but due to the fact that Dr. Hassibi is, unfortunately, such a busy person, it is not viable. My solution is to have a graduate student or equivalent guide us because it gives him teaching experience and us more time to learn from the best. Next semester, since this is called the "Caltech STEM Research Program", I expect more of the research facet to be implemented. I hear that other branches, such as the Chem group, have published a paper. That experience of researching unknown territory was what intrigued me because I think that is what I want to pursue in the future. I also expect Navid to spend more time to help us since the current trend is doing nothing.

An image Matthew showed what class should look like. I agree

       As we roll into our independent projects, I will explore Machine Learning applied to hand gestures and applications that benefit from it. In the meantime, I will touch upon my linear algebra during the winter break since it is too late to go over it during semester 2.

Hand recognition program that I will attempt to understand and program myself