Curing Diseases

Misdiagnosis is a very common problem with these diseases, causing officials to offer the wrong treatment. To prevent this, we have used datasets from Kaggle to deal with a variety of symptoms.

Comment

Inspiration

Misdiagnosis leads to improper treatment and delayed car. My teammate's grandma in India had a similar issue. This is often known as medical malpractice. Often times, this doesn't happen because of negligence of a doctor. It can be due to errors in scans and various symptoms that combined cause officials to identify the wrong treatment.

What it does

It's a fully functioning classification model with a high accuracy. The goal is to use the symptoms to classify it to the right disease to solve misdiagnosis problems. We collect data based on a diverse symptoms that patients report of various demographics. The 3 models that we have used are important because it separates various diseases and treatments by race, a factor often ignored. We have used some special methods like SHAP to successfully make our predictions understood by professionals that eventually use it. Gender and economic status are also important factors in healthcare that can cause the wrong treatment to be identified.

How we built it

Datasets from kaggle were the backbone of this project. 3 valuable datasets were able to help us identify the health issue. There are various factors we worked with such as CT scans, rage, gender, status, skin, etc. The decision trees were able to split each individual data into branches. We were able to prioritize what feature is most important for a certain specific disease.

Challenges we ran into

Our model failed to run many times initially. However, we had to find the right dataset that could be supported by our model. Our model was pretty small, but it was fed a lot of data from Kaggle which is what allowed it to be able to detect symptoms and classify it based on the right problem. This problem was about right classification.

Accomplishments that we're proud of

We're proud of being able to make a model for a variety of diseases. Developing a fully functional classification model with a high accuracy rate in a short period of time.

What we learned

We expanded our knowledge in developing classification models and also deepened our understanding in the different diseases and their effects on people and how our models can help predict and cure the patients before it gets worse.

What's next for Curing Diseases

We hope to be able to make models like this and expand the symptoms we have so we an be able to detect any type of disease using all the symptoms that exist.

Built With

Share this project:

Updates