How bhagwati2316/titanic-passenger-survival-chance-using-deep-learning works
This small Streamlit application collects Titanic passenger attributes, encodes categorical fields and scales numeric ones, then runs a pre-trained ANN to produce a survival probability and a threshold-based message. The model and preprocessing encoders are repository artifacts; the app loads them at runtime rather than training them. The sampled app.py shows the full workflow, so no separate service or storage layer is indicated.
Main entry points: app.py Feature Preparation · model.h5 Survival ANN · label_encoder.pkl Sex Encoder · onehot_encoder.pkl Embarked Encoder
Components 9
Other components
- Passenger
- Passenger Inputs
app.py - Feature Preparation
app.py - Sex Encoder
label_encoder.pkl - Embarked Encoder
onehot_encoder.pkl - Feature Scaler
scaler.pkl - Survival ANN
model.h5 - Survival Probability
app.py - Survival Result
app.py
Connections 8
- Passenger → Passenger Inputs: enters attributes
app.py - Passenger Inputs → Feature Preparation: builds dataframe
app.py - Feature Preparation → Sex Encoder: encodes sex
app.py - Feature Preparation → Embarked Encoder: encodes port
app.py - Feature Preparation → Feature Scaler: scales fields
app.py - Feature Preparation → Survival ANN: predicts from features
app.py - Survival ANN → Survival Probability: returns probability
app.py - Survival Probability → Survival Result: thresholds and displays
app.py