rdjverse/CancerGuardian
CancerGuardian is a machine learning-powered web app that helps predict breast cancer diagnoses based on cytology measurements. 🩺✨ Built with Streamlit, Scikit-Learn, and Plotly, this tool visualizes tumor characteristics and provides predictions using a trained model. 🚀
What's novel
CancerGuardian is a machine learning-powered web app that helps predict breast cancer diagnoses based on cytology measurements. 🩺✨ Built with Streamlit, Scikit-Learn, and Plotly, this tool visualizes tumor characteristics and provides predictions using a trained model. 🚀
Code Analysis
6 files read · 3 roundsA Streamlit web app that uses a scikit-learn LogisticRegression model on the Wisconsin Breast Cancer dataset to predict benign vs. malignant classification from 30 cytology measurements, with a Plotly radar chart visualization.
Strengths
The app is functional and well-organized for its scope, with clear separation between training and inference, and the radar chart provides a reasonable visual aid for comparing tumor characteristics across mean/SE/worst dimensions.
Weaknesses
The README falsely claims 'deep learning techniques' when the model is a basic LogisticRegression, the requirements.txt has a typo (scikit_learn vs scikit-learn) that breaks installation, there are zero tests, no error handling for missing files or invalid inputs, and the entire project is a textbook tutorial-level exercise on a well-known public dataset with no novel contribution.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
5
Contributors
2
Files
12
Active weeks
2
Repository
Language
Python
Stars
1
Forks
0
License
MIT