Hello everyone,
Now that I am over half way done with Fall term 2024, I figured I would check in. So far my project in CS461 is going very well. Me and a small team are working on using AI to help my local public transit system improve their services. This is honestly a really fascinating project because it has potential to make real impact in people’s lives.
So far, our team has worked together very well to solve problems. Much of our work at this point has been getting familiar with some of the technologies we will be using, like Malloy and BigQuery. We have also spent time thinking about how we will conduct our analysis on the data we receive. For a brief overview, our team will be trying to improve rider satisfaction and safety, and to provide predictive maintenance warnings to the transit company. We are going to be analyzing rider feedback, emergency calls, weather, GPS, and maintenance data to make our predictions. With that, what I love about this project is the shear amount of directions we can take once we begin training on the data. We may perform sentiment analysis, feature selection, dimension reductions, anomaly detection, game theory, etc. I think it is this aspect that makes the project so fascinating.
Our last hurdle in our project is obtaining the data. We have not obtained all the data but are in the process of doing so. We are confident it will happen within a month.
On another yet similar note, CS461 is going really well. I was a bit skeptical of the amount of writing assignments present in the course, but I have grown to appreciate them, as they have forced our team to do preemptive research even though we don’t have the data. One thing that I think would be helpful for future CS461 students is that research projects have a larger focus on literature reviews. I think being concerned with literature reviews is more important than having a perfect plan this early in the game.
As a side-note for you Python users, I have been using Google Calob for a lot of my course work this term, and I must say: Wow! Google Colab really is an amazing technology for machine learning and mathematics. If you find yourself doing any programming in Python that is very segmented, meaning the codebase is largely unrelated between sections, strongly consider Colab. It really is great.
Finally, I will leave you all with a word of advice: Drink more water, read more books, and sleep more.
Brayden Edwards
Leave a Reply