#!/usr/bin/env python
# coding: utf-8
#
Build a SVM modle for Face Recognition Problem
# ---
#
# We will use a very famous dataset, called Labelled Faces in the Wild, which
# consists of 1288 faces of famous people, and it is available at http://viswww.cs.umass.edu/lfw/lfw-funneled.tgz.
#
# However, note that it can be easily imported via scikit-learn from the datasets class.
# Each image consists of 1850 features: we could proceed by simply using each of them in the model.
#
#
#
# Fitting a SVM to non-linear data using the Kernel Trick produces non- linear decision boundaries.
# In particular, we seek to:
# * Build SVM model with radial basis function (RBF) kernel
# * Use a grid search cross-validation to explore ran- dom combinations of parameters.