🧠 ML Model Benchmark — Olivetti Faces (sklearn)

📋 Overview

This project tests and compares multiple machine learning and deep learning models on the Olivetti Faces dataset from sklearn.datasets.fetch_olivetti_faces. It is designed to evaluate model performance in terms of accuracy and training speed for a facial recognition task.

🎓 Originally built for the CS42 Model Selection unit. The notebook walks through training, evaluation, side-by-side plots, and a small interactive UI for testing the best model.

---

🧩 Models Tested

Category	Model	Library
Classical ML	K-Nearest Neighbors (K=3)	scikit-learn
Classical ML	Decision Tree (max_depth=25)	scikit-learn
Ensemble ML	Random Forest (100 estimators)	scikit-learn
Neural Network	Multi-Layer Perceptron (MLP, 256 hidden units)	scikit-learn
Gradient Boosting	XGBoost	xgboost
Deep Learning	Convolutional Neural Network (CNN, 100 epochs)	TensorFlow / Keras

---

🎯 What It Measures

Metric	Description
Training Accuracy	How well the model fits the training data
Test Accuracy	How well the model generalizes to unseen data
Training Time (s)	Time required to train each model

---

🧠 Dataset

Property	Details
Name	Olivetti Faces (fetch_olivetti_faces)
Type	64×64 grayscale human face images
Classes	40 subjects
Samples	400 images total
Task	Facial recognition (multi-class classification)

---

🖼 Samples & Charts

Dataset preview:

MLP model predictions:

Compare training accuracy:

Compare test accuracy:

Compare training time (log-scale):

Interactive UI for model testing:

---

⚙️ How to Run

# 1) Create and activate a virtual environment (optional but recommended)
python -m venv .venv
source .venv/bin/activate          # On Windows: .venv\Scripts\activate

# 2) Install dependencies
pip install -r requirements.txt

# 3) Open the notebook
jupyter notebook "ML models comparsion_v2.ipynb"

💡 Run on Google Colab for free GPU acceleration on the CNN — just upload the notebook to colab.google.com.

---

🧰 Requirements

Package	Why
Python ≥ 3.10	Required runtime
numpy, pandas	Data handling
matplotlib	Plotting comparison charts
scikit-learn	Olivetti dataset + KNN, DT, RF, MLP
xgboost	Gradient boosting baseline
tensorflow	CNN training
ipywidgets	Interactive UI inside the notebook

Install everything with pip install -r requirements.txt.

---

📁 Project Structure

ML-Models-Comparison---Speed-Accuracy/
├── ML models comparsion_v2.ipynb   # Main notebook (training + comparison + UI)
├── requirements.txt                # Python dependencies
├── LICENSE
└── README.md

---

🎓 Learning Objectives

1. Understand the accuracy / speed trade-off across model families
2. Compare classical ML vs boosting vs deep learning on the same data
3. Read training-curve plots and interpret signs of overfitting
4. Use ipywidgets to build a quick interactive demo for any classifier

---

🛑 Troubleshooting

Issue	Fix
ModuleNotFoundError: tensorflow	Run pip install tensorflow (CPU build is fine for 64×64 images)
First-run download fails	fetch_olivetti_faces() downloads ~1 MB to ~/scikit_learn_data/; check your network
Notebook UI widgets blank	pip install ipywidgets && jupyter nbextension enable --py widgetsnbextension

---

📄 License

MIT — Credits to CS42.org.