I was inspired by the idea that a single person interested in technology can be able to create a project that can change the lives of others and the people around them in a positive way. I wanted to contribute to this community by creating a technology which anyone could use to possibly save their life.
What it does
This model uses Machine Learning to assess and predict whether or not a patient has lung cancer based on their Gender, Age, Smoking, Anxiety, Chronic Disease, Alcohol Consumption, and Shortness of Breath. Each of these inputs are used to build a pattern using Machine Learning and basic data science to see patterns between these characteristics and the probability of having lung cancer.
How I built it
I used a dataset on Kaggle and used python, PyTorch, skit learning, and Matlab to plot, predict, and build the model.
Challenges we ran into
The most common challenge was trying to reduced the Mean Absolute Error as much as possible, by making it closer to zero. Another challenge I had come across multiple times was the fact that I wasn't overfitting or undercutting any of the data, to make sure that I get the optimal solution, but to also make sure that the model, after going through the training dataset and validation set, would be able to perform as it was intended to outside the validation set and work on other sets of data that it hasn't encountered before.
Accomplishments that we're proud of
I'm proud that this was my first Hackathon, and that I was able to show off my skills in Data Science and Machine Learning in a competitive setting.
What we learned
I learned a lot about over and underfitting, making sure my Machine Learning algorithm works as intended.
What's next for Lung Cancer Prediction Model
I want to be able to extend this program by having it use edge detection to look at pictures and determine whether or not a person has lung cancer.