MULTI-DISEASE CLASSIFICATION

Inspiration

Our team was interested in the Multi-Disease Classification project from Phyla. We thought we can use Python to solve this problem.

What it does

The program reads in the dataset and applies it to multi-label classification algorithms such as problem transformation algorithms and a adapted algorithm

How we built it

The dataset was pre-processed to separate the different Disease labels so that test and trained data could be applied to multi-label classification algorithms.

Challenges we ran into

The challenges we had were:

1. The team is relatively new to machine learning so it took us longer to figure out what algorithms we needed and how to implement them
2. We were unable to apply the cohen's kappa scoring within the time allotted.

Accomplishments that we're proud of

1. We were able to split the labels into their own columns to create the multi-label dataset to apply the classifications used
2. We did get a good result when we applied the adapted algorithm to the dataset.

What we learned

1. We learned about multi-label classification which the team did not have much experience with previously
2. We learned about a new scoring method for algorithms: Cohen's kappa.

What's next for MULTI-DISEASE CLASSIFICATION

With more time, we would like to build visualizations and explore more precise ML.

Built With

• python