Identifying Malignant Tumors with Machine Learning
About this project
Using data centered around the measurement of a variety of tumors, we created 3 machine learning models and compared the performance of each in classifying tumors as malignant or benign. We catered our approach, explanation, and demo towards medical professionals and explained a variety of model performance metrics, as misdiagnosing a tumor could have dramatic consequences. Due to this, we prioritized sensitivity over accuracy as we do not wish for a malignant tumor to go undetected while also considering specificity in order to avoid in unnecessary invasive surgery.
More projects by Brianna King
Using R Shiny to Inform Community Officials
Using a leaflet interactive map, we created a multilevel, informational dashboard that enables...
Domesticated Animal Image Classification with Python
In this use case, my colleagues and I collaborated on creating and optimizing a convolutional...