Project Data Science by Jasmine Hale

Machine Learning

jasminehale — 2025-05-14 20:40:25 UTC

Machine Learning is a factor of artificial intelligence that allows computers to learn from data without being told exactly what to do. When we started using technology in this class we coded the computer to do exactly what we wanted. However now we are experimenting with ways to let the technology do what we want it to based of examples we give it. When you think of this you can think of Fraud detection. Computers look for fraud without constantly being told to do so. Fraud detection identifies fraud transactions in banking and finance so this perfectly shows an example of Machine learning.

Filtering Methods

jasminehale — 2025-05-15 17:12:24 UTC

Content-based and collaborative filtering are two main factors we have focused on. These filtering methods are exactly what we use for machine learning. But specifically Content-based filtering is a recommendation system that suggests items similar to what a user has liked in the past and Collaborative filtering is a technique used in recommender systems to predict what a user might like based on what other similar users have liked or rated.

Rating our songs

jasminehale — 2025-05-15 17:26:10 UTC

We rated songs so that we could use them to explore content based and collaborative filtering in machine learning. We will use these ratings for the computer to see how we like each song. The computer ended up telling us their content based estimate based on what attributes we like about the song providing their overall guess to our rating of the song. Our ratings even went as far in collaborative filtering when it gave us our prediction based on our classmates! Our ratings went far providing us with tons of information.

Complexity and Attributes

jasminehale — 2025-05-16 07:52:00 UTC

My top three attributes were Dance ability, speechiness, and acousticness. These attributes are things I like most about a song. This matters because now the computer can take these songs and find other ones I may like. This helps us learn more about machine learning because it gives the computer more about our preferences and so it can do more for us. Choosing our level of complexity was important so that we could learn farther about our information. Overall these are things that help us study machine learning.

Content-Based Filtering Models

jasminehale — 2025-05-16 08:11:26 UTC

Content Based filtering once again is having the computer recommend things based on what users have liked in the past. We used songs and their attributes to our liking to predict how we would rate our songs. At first we used one attribute to create a model. My first attribute model was slightly inaccurate to what I would actually rate my song. However when we started adding more attributes it made my models more exact. My second attribute model already started showing that increase in accuracy. The content based data was officially gathered when we figured out that when you add more attributes more accuracy the model presents. The same thing happened with collaborative filtering. The more attributes we added the more accuracy we found. However since this is a different type of machine learning it gives us even more data. It gives us people that we actually would have ratings in common with and it let us know who we have similar and different tastes than. These findings add to my understanding of collaborative filtering. My only surprise was that I actually don't have a favorite method. I think they are both great ways to gather data. I think the only way this study to get even more accurate is if I had a better selection of attributes to choose from. If I liked the options more maybe it would be even closer to what I actually like.

Data Ethics of Machine Learning

jasminehale — 2025-05-16 08:16:36 UTC

These Ethics have a lot to do with bias. People believe that since it is a computer it could make errors. When we see skewed numbers or unreasonable answers we can be aware of these and find out how to not make these biases happen again. The most important information for the general public to know about machine learning is that it only is trying to benefit us. The information it gives us is just trying to give us recommendations that we love. It is how machinery works and how we get it to work. The public should see the cool things we found out so they can see why people should learn about the great things that machine learning can do.