💎 Diamond Price Prediction

A machine learning project that predicts diamond prices from a handful of physical measurements (carat, length, width, depth). It trains and compares four regression models — Linear Regression, Support Vector Regression, Extra Trees, and Gradient Boosting, and serves live predictions through a Streamlit app.

Requirements

Python 3.9+ is recommended. Install the packages used across the project:

pip install streamlit pandas numpy scikit-learn joblib matplotlib seaborn

Running the Streamlit app

From the project root:

streamlit run app.py

---

Models included

Model	Training script	Saved as
Linear Regression	linear_regression.py	lr_model.joblib
Support Vector Regression	support_vector_regression.py	svr_model.joblib
Extra Trees Regression	extra_trees_regression.py	etr_model.joblib
Gradient Boosting Regression	gradient_boosting_regression.py	gbr_model.joblib

Each tune_*.py script sweeps hyperparameters for its model, prints results, and saves CSVs + PNG charts to output/. The FINAL_PARAMS in each training script reflect the best values found during tuning.

Project structure

project-root/
├── app.py                          # Streamlit app — run from here
├── data/
│   └── Diamonds Prices2022.csv     # raw dataset
├── output/                         # auto-created by tune_*.py scripts
│   ├── linear_regression/
│   ├── support_vector_regression/
│   ├── extra_trees_regression/
│   └── gradient_boosting_regression/
├── src/
│   ├── preprocessing/
│   │   └── preprocessing.py        # shared load_test_split() / load_full_split()
│   ├── models/                     # models with feature selection
│   │   ├── linear_regression.py
│   │   ├── support_vector_regression.py
│   │   ├── extra_trees_regression.py
│   │   ├── gradient_boosting_regression.py
│   │   └── hybrid_model.py
│   ├── full_models/                # models without feature selection
│   │   ├── full_linear_regression.py
│   │   ├── full_support_vector_regression.py
│   │   ├── full_extra_trees_regression.py
│   │   └── full_gradient_boosting_regression.py
│   ├── tuning/                     # tuning files to find best hyperparameters
│   │   ├── tune_linear_regression.py
│   │   ├── tune_support_vector_regression.py
│   │   ├── tune_extra_tree_regression.py
│   │   └── tune_gradient_boosting_regression.py
│   └── saved_models/               # auto-created — holds .joblib files + metrics.json
│       ├── lr_model.joblib
│       ├── svr_model.joblib
│       ├── etr_model.joblib
│       ├── gbr_model.joblib
│       └── metrics.json
└── README.md