Day 9: Visualization and AI Overview

Data Visualization: McDonalds Dashboard

I learned how to create a simple dashboard that consisted of sales, charts, and customer satisfactions. With this visualization, users could easily navigate between pages, see raw data compared to visualized data, and have easy access to contact information. With this dashboard data can be more easily understood and more easily analyzed whether it's for company improvement or company stats.

What is AI?

AI works based on data that it is taught whether it is text from a news outlet or a collection of art pieces. With this data, it can use its taught information to provide responses to prompts or stitch together words or images that correlate with the prompt. This technology can be used to generate all sorts of media whether it's a story, a summary, an image, or art.

How Humans Determined Who Would Survive vs Code:

Similarly to AI, we were given a set of videos to be our clues/our human "data." Based on these videos, we gathered that children and women were typically prioritized over men, so based on the people's names and ages, we guessed that all women and children survived, with one or two randomly selected men. My code assigned woman survivors as well and tested it along real survivors. While it could be better, this process had ~70% accuracy, which was not bad for testing only one condition.

What I Learned:

Besides the AI Titanic code, I also learned about some AI ethics in STEM in Society today as well. Specifically whether or not AI developers should have free reign to certain data, or if data should be heavily restricted. I found this discussion interesting along with the way it connected to our data collection process in STEM Skills.

Comments

QR Code

QR Code to Katie's Blog

Day 7: Are Animals People?

If you do not meet all the criteria for personhood, does that mean you are not a person? A person could be capable of a certain criteria more than another criteria and still be a person. If a person lacks a certain criteria like morality, they could still be highly sentient or highly reasonable. I think 'matching the criteria' is a spectrum. For example, a being that we consider would normally consider capable of personhood (such as a human) can be born uncapable or weaker in a certain criteria, but that does not inherently make them not a person. The idea of these criteria is a blanket term, and the term 'personhood' itself is broad and nonrestrictive.

Day 11: Dr. Schuman

Guest Speaker: Dr. Schuman Today we got firsthand experience in creating data for a machine learning activity. We visited a site where we were able to feed information through series of pictures of either a facial expression, a hand signal, or certain objects. With this information, our machine was able to learn what traits were associated with certain items through an accumulation of mathematical formulas. Additionally we also got to play with an app that learned from other people's drawings and was able to complete our drawings and portray certain prompts (such as dog, cat, mermaid, lighthouse etc). Overall it was an engaging experience that allowed us to learn about machine learning data and experience collecting data ourselves. Pictured: Training Data of My Classmate

GSSE Blog Daily: Katie Lam

Search This Blog