gistfile1.txt

{
 "metadata": {
  "name": "Untitled6"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "!wget http://deeplearning.net/data/mnist/mnist.pkl.gz",
     "language": "python",
     "metadata": {},
     "outputs": []
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "import cPickle,gzip\nfrom sklearn.ensemble import RandomTreesEmbedding\nfrom sklearn.svm import LinearSVC",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "f = gzip.open('mnist.pkl.gz', 'rb')\nmnist_train_set, mnist_valid_set, mnist_test_set = cPickle.load(f)\nf.close()",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 2
    },
    {
     "cell_type": "heading",
     "level": 4,
     "metadata": {},
     "source": "Training with random data"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "encoder = RandomTreesEmbedding(max_depth=None,n_estimators=100,n_jobs=-1)\nencoder.fit(randn(1000,784))",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 3,
       "text": "RandomTreesEmbedding(max_depth=None, min_density=None, min_samples_leaf=1,\n           min_samples_split=2, n_estimators=100, n_jobs=-1,\n           random_state=None, verbose=0)"
      }
     ],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "X = encoder.transform(mnist_train_set[0])\ny = mnist_train_set[1]",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "%timeit cls = LinearSVC() ; cls.fit(X,y)",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": "1 loops, best of 3: 34.2 s per loop\n"
      }
     ],
     "prompt_number": 14
    },
    {
     "cell_type": "heading",
     "level": 4,
     "metadata": {},
     "source": "Training with original data"
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "encoder = RandomTreesEmbedding(max_depth=None,n_estimators=100,n_jobs=-1)\nencoder.fit(mnist_train_set[0][:1000,:])",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 22,
       "text": "RandomTreesEmbedding(max_depth=None, min_density=None, min_samples_leaf=1,\n           min_samples_split=2, n_estimators=100, n_jobs=-1,\n           random_state=None, verbose=0)"
      }
     ],
     "prompt_number": 22
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "X = encoder.transform(mnist_train_set[0])\ny = mnist_train_set[1]",
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 23
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "%timeit cls = LinearSVC() ; cls.fit(X,y)",
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": "1 loops, best of 3: 3.33 s per loop\n"
      }
     ],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": "",
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}