Unit 7 Data Science Project- Milena Flynn by Milena Flynn

What is Machine learning!?

milenaflynn — 2025-05-14 19:53:32 UTC

Machine learning is teaching computers to learn from data, like how we learn from experience. Instead of telling a computer exactly what to do, we give it lots of examples, and it figures out the rules on its own. For example, facial recognition. On smart phones with facial recognition the phone learns to be able to recognize you by our face as it remembers your face shape, eyes, nose, etc.

Filtering methods ( content based)

milenaflynn — 2025-05-14 19:55:48 UTC

CONTENT BASED- Content-based filtering is a method used by recommendation systems to suggest items to users based on their past preferences. It works by analyzing the characteristics of the times the user has interacted with, and then will recommend similar items.

Filtering methods ( collaborative)

milenaflynn — 2025-05-14 20:15:56 UTC

COLLABORATIVE- Collaborative filtering is a recommendation method that suggests items to users based on the preferences of similar users. It identifies users who have similar tastes and recommends items that those users have liked or interacted with. For example, if many users with a similar movie preferences to yours enjoyed a particular film, the system will recommend that film to you.

Collecting our data

milenaflynn — 2025-05-14 20:22:27 UTC

The purpose of listening to the 23 songs and ranking them on a scale of 1-10 on how much we liked them was so that the computer was able to get an idea of what kind of music we liked. The Attributes that I chose were dance ability, speechiness, and durationMs. I chose these three because they worked best for the songs that I liked meaning on the graph they had the strongest correlation and the line of best fit.

Choosing your level of complexity

milenaflynn — 2025-05-15 17:13:46 UTC

Machine learning is made to ensure that the model works well with data that is unseen. This means that the model has to balance the training data and the test data.

Data Ethics Of Machine Learning

milenaflynn — 2025-05-15 17:21:00 UTC

Machine learning is a type of artificial intelligence that allows computers to learn from data without being programmed to it exactly. Instead of following preset rules, these systems identify patterns and make predictions based on the data that they were fed.

Its important to know that:

Machine learning is all around us in the sense that we use it daily without us even knowing.

It's not always perfect as machine learning is a rapidly developing field so new techniques and skills are still evolving.

Is there any bias?

Yes, theirs many different forms of biases that can happen with the methods such as..

data bias

Sampling bias

Algorithmic bias

Conformation bias

Measurement bias

Evaluation bias

etc..

Being aware of these potential biases allows developers and users to take steps to cancel them out and create fairer, more accurate machine learning models.

Collaborative filtering ( 7.8-Colab 5)

milenaflynn — 2025-05-16 18:38:30 UTC

The classmate that has the most similar ratings to me was Jaelyn Forgey. For the ratings its all pretty accurate but only a few of the songs are off. To make it more accurate I should Have been more diverse with the ratings and spread out the numbers so they weren't all the same.

Content Based Filtering Models (7.6- Colab 4)

milenaflynn — 2025-05-16 18:41:26 UTC

For the one attribute model The predictions were pretty accurate as on one song I had rated it a 6 and the computer had made a prediction saying I would rate it 6.52. Then with the two attribute model It was still pretty accurate but the rating was a little more off than with the one attribute model. My guess to why that happened would be because of the song and maybe one attribute fit the one song better than the other.