AI Week 2: Coding Netflix?

We survived the first week of AI online but the second week felt much more daunting and I was honestly starting to feel pretty overwhelmed. My favorite thing about Computer Science on the block plan is how collaborative it is. We, as a class, are constantly working together to understand new material and finish our assignments. I have very rarely done the majority of my homework locked in a room by myself, I am always finding a friend in my class to work with or just working in the lounge of our coveted Tutt Science. With all this being said, I was getting frustrated by the lack of collaboration on assignments. 

I was incredibly thankful to have a friend in the class who I am often texting about class or assignments, and one night we sat on zoom for a solid 4 hours to pair program (programming together and bouncing ideas off one another). With this week being overwhelming as class picked up pace we also got a really fun and challenging assignment.

We were coding Netflix’s recommendation engine. As some people may or may not remember, Netflix started out simply by the user being able to order a movie to their house and then mail it back when they were done watching. I remember the days of waiting for the next Netflix movie to arrive. However, Netflix knew that if they could nail down their recommendation engine it would be incredibly profitable. After trying simply within their company they decided to release a lot of their data and start a competition. The first person to get it would win one million dollars. After much time and many being unsuccessful they finally made smaller prizes of 10,000 dollars for anyone who could make progress they deemed significant and post a research paper explaining how they did it. Long story short, the code that Netflix deemed most useful was from a guy who programmed in his college dorm.

We took ideas we had learned the week prior, with the nearest neighbor algorithm and worked to code the Netflix recommendation engine. Of course we did not use as much data as Netflix, but I would say that in the end it was a success story. I found myself really excited by this project which made me mourn not getting to be on campus. I am still grateful for zoom calls with my professor and classmates who are willing to work together despite being miles apart.

Artificial Intelligence Online: Week 1

Artificial Intelligence sounds scary in itself, but when you add on the online portion it sounds far too intimidating. While I was hesitant to take this class once it was moved to online, I decided to continue with it because it was something I had been looking forward to learning all year. The first day of class my professor announced that he knew this would be a lot of learning and adjusting as we go, everyone would be adjusting to online learning and Richard was willing to be flexible with us. In normal block plan fashion we dove into topics of AI on day 1. 

Our first task being to define, What is Artificial Intelligence? As we bounced around ideas we landed on AI being a classification problem, where one is taking labeled data and learning from examples. AI is a statistical analysis of data. For this class, we are focusing on machine learning, which goes hand in hand with AI. This class seems to be working with a lot of different topics including algorithms, linear algebra, and data science. The first week we worked a lot with implementing well known machine learning algorithms such as Nearest Neighbor Model. This algorithm allows us to plot different points on a graph based on our labeled data and ask the computer to find the nearest neighbor to a specific point. As you may be able to assume, this takes a lot of both math and computer science so a good portion of our lectures were simply making sure we were comprehending what mathematics we were going to have to implement. We work really hard in this class to understand the “why” and “how” of our problem before trying to start coding. 

While I will be the first to admit I am not a fan of distance learning I am a fan of having recorded lectures. I have found myself motivated (or confused?) enough to go through and watch my lectures again, to pause and take more detailed notes, and to keep track of things that I may have more questions about.