How mahammedsathyala/house-price-prediction-ibm-skillsbuild works
This repository is a King County house-price analytics and regression project. Its main workflow profiles and cleans transaction data, explores price patterns, trains and evaluates regression models, then serves interactive price estimates and analytical views through Streamlit. The graph also shows the model and report artifacts shared by training and the app. The app excerpt is partial, so its documented dashboard, EDA gallery, and insights are represented without assuming unshown internal call paths.
Main entry points: streamlit_app.py Streamlit application · model_training.py Model training and evaluation · main.py Training orchestrator · data_cleaning.py Cleaning and features
Components 13
Data analysis
- Housing data
data/data.csv - Data profiling
src/data_understanding.py - Cleaning and features
src/data_cleaning.py - EDA charts
src/eda.py
Modeling
- Training orchestrator
main.py - Model training and evaluation
src/model_training.py
Interactive application
- Streamlit application
app/streamlit_app.py
Runtime artifacts
- Best model
models/best_model.pkl - Feature columns
models/feature_columns.json - Evaluation results
models/eval_results.json - Feature importance
models/feature_importance.json - Generated charts
visualizations
Other components
- Home seeker
Connections 19
- Home seeker → Streamlit application: uses
- Training orchestrator → Housing data: profiles
- Training orchestrator → Data profiling: calls
- Training orchestrator → Cleaning and features: calls
- Cleaning and features → Housing data: reads
- Training orchestrator → EDA charts: calls
- Training orchestrator → Model training and evaluation: calls
- Model training and evaluation → Best model: saves
- Model training and evaluation → Feature columns: saves
- Model training and evaluation → Evaluation results: saves
- Model training and evaluation → Feature importance: saves
- Model training and evaluation → Generated charts: writes
- EDA charts → Generated charts: writes
- Streamlit application → Cleaning and features: loads data
- Streamlit application → Best model: loads and predicts
- Streamlit application → Feature columns: loads
- Streamlit application → Evaluation results: loads
- Streamlit application → Feature importance: loads
- Streamlit application → Generated charts: displays