Blog Three

02. The Final Chapter

Content

Going into this project, I expected there to be disparities based on age, sex, and race. However, I greatly underestimated the disparities; for example, females of Indian descent (regardless of age) are 14.6218 times more likely to be falsely misidentified than males of Indian descent. I also expected that there would be improvement between all three versions, but was surprised to see that there was a slight decline in accuracy from version 11 to 13. Overall, I learned that women, older people, and people of color are all unfairly associated in rank_one. The next steps for this project is to look at the rank_one algorithm itself, and see what causes this bias. By noting which demographics are the most unfair, the developers can aim to mitigate this issue in their programs.

Product

One thing that I learned from my final presentation is to talk slow. I tend to speed up my words when I'm nervous but talking slow helps me sound more presentable and also calms me down. While practicing for my presentation and oral defense, I learned sometimes its good to just take a deep breath to recollect my thoughts. Doing this in my presentation definitely helped me be more in control of my public speaking. One thing that I learned from my paper is that peer reviewing is really helpful. Especially getting someone who hasn't heard of my project was beneficial, as I could see which parts didn't make sense and needed more clarification. No matter the importance of the writing, it is always a good idea to review it with one or two outside opinions before submitting it.

Process

One of the biggest things that I have learned about myself as a student is that I need to have deadlines, whether created by myself or others, in order to stay on track for projects. Breaking my project down into smaller steps made it much easier for me to follow on schedule (even if it meant dedicating hours on end to meet those created deadlines). In the future, I will implement this in college by setting myself deadlines for smaller steps in big projects so I don't end up doing everything at the end. Another thing that I learned about myself is that even though I get distracted when working with friends, I did have some of my best cram sessions with them. In college, I definitely do want to work on assignments with friends again, but will work on being more focused during these sessions.

Acknowledgements & Future

I want to thank Mrs. Dobos and Mr. McBride for their invaluable guidance and support throughout my project. Their assistance enabled me to effectively manage my work. By providing deadlines and check ins, Mrs. Dobos and Mr. McBride ensured that my project reached its fullest potential. I also want to thank everyone who attended my presentation - Mrs. Bukowski, Mr. Solis, Mrs. Kerr. I am so glad that you could see the compilation of all my hard work this year!

While this chapter is over, I will be starting my next one at UCSD! I'll be studying Mathematics with Computer Science and will hopefully get to learn more about AI and facial recognition!