Inspiration
I got the idea for the project when I completed my first of 3 ML specialization courses made by Andrew Ng. I figured this was not only a great way to end of a month of coding and learning of basic ML but a way to prove myself. Sadly I wasn't able to get to it because of school, however I found out about MLH weekend hackathons and decided to make it for this weekend!
What it does
My project incorporates many math concepts (Linear Algebra and Calculus in specific) and creates a binary classifier using some common algorithms in the realm of Supervised ML.
How we built it
I used Python's core ML libraries (pandas, numpy, matplotlib, and scikit-learn) in order to create my project.
Challenges we ran into
The biggest challenge I encountered was the many, MANY overflow errors I was receiving. (It took me a few hours to realize that all I needed was a bit of numpy code to fix it np.clip.
Accomplishments that we're proud of
The main accomplishment I am proud of was the ability for me to even participate in my first hackathon, as well as finish this project.
What we learned
I learned that I should adapt myself even more with "the tools of the trade" as I was struggling for many many hours just to debug something that could be solved with a line of code. I want to adapt myself more to these projects to prevent these type of slow-downs in upcoming, and bigger projects.
What's next for Breast Cancer Classification
There are an extensive amount of models that can be used for the task of not only binary classification, but also multi-class classification. However for binary classification, I want to attempt to use random forest and deep learning models to see how projects with these types of models are like.
Log in or sign up for Devpost to join the conversation.