Now Arriving: New Display Board

This past week I worked on my display board, original work speech, and toxicity model. The display board took me a while, but in the end, I’m happy with how it looks. Having a plan in place definitely helped, and so did having past examples to look at. I hope that I get to display it at Research Showcase sooner rather than later. Meanwhile, I’ve been expanding on my original work speech. Because I’ve been simultaneously working on my DECA presentation, it feels like I’ve had a lot of speaking to work on, but ISM speeches are easier because I know my topic more intimately. For the toxicity machine learning model that I have been working on, I managed to get a mean absolute error that was close to the benchmark using a rather simple method. However, I am now trying a new method, which has come with some obstacles. For one, I want to try using Amazon’s AutoGluon, but I’m not as familiar with it, which makes it harder to train with ECFPs or even tabular data as I’ve been doing before. With some time, I’ll figure it out, and I’m glad that I am already at a pretty good starting point with the simple, fast-training model. This week, I’m excited to be off school for two days and have an etiquette lunch!