How syed-md-abdud-dayan/pneumonia-detection works
This repository documents a notebook-based chest X-ray classification workflow: load and inspect image data, preprocess and augment it, train a VGG16-based binary classifier, then evaluate and report results on held-out test data. The README describes the workflow and model, but no notebook source was available to verify internal calls or implementation details; the edges below therefore reflect only documented data flow and operations.
Main entry points: Pneumonia Detection.ipynb Pneumonia workflow · Pneumonia_Detection_Images Saved visualizations
Components 16
Data analysis
- Dataset inspection
Pneumonia Detection.ipynb - Class and image EDA
Pneumonia Detection.ipynb
Image preparation
- Image preprocessing
Pneumonia Detection.ipynb - Training augmentation
Pneumonia Detection.ipynb - Class weighting
Pneumonia Detection.ipynb
Model training
- VGG16 classifier
Pneumonia Detection.ipynb - Model training
Pneumonia Detection.ipynb - Training callbacks
Pneumonia Detection.ipynb - Best model
Pneumonia Detection.ipynb
Evaluation
- Independent test set
Pneumonia Detection.ipynb - Evaluation metrics
Pneumonia Detection.ipynb - Evaluation reports
Pneumonia Detection.ipynb - Saved visualizations
Pneumonia_Detection_Images
Other components
- Notebook user
- Google Drive data
- Pneumonia workflow
Pneumonia Detection.ipynb
Connections 17
- Notebook user → Pneumonia workflow: runs
- Pneumonia workflow → Google Drive data: loads data
- Pneumonia workflow → Dataset inspection: inspects
- Pneumonia workflow → Class and image EDA: analyzes
- Google Drive data → Image preprocessing: provides images
- Image preprocessing → Training augmentation: prepares training
- Image preprocessing → Model training: provides inputs
- Pneumonia workflow → Class weighting: calculates
- Class weighting → Model training: weights classes
- Pneumonia workflow → VGG16 classifier: builds
- VGG16 classifier → Model training: provides model
- Model training → Training callbacks: uses
- Model training → Best model: saves best
- Best model → Independent test set: evaluated on
- Independent test set → Evaluation metrics: supplies predictions
- Evaluation metrics → Evaluation reports: summarized as
- Pneumonia workflow → Saved visualizations: produces plots