Mean-Variance.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# SGD"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k3 = [\n",
    "    0.927102805521,\n",
    "    0.934579440589, \n",
    "    0.902621723293]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k3_lr2 = [\n",
    "    0.530841121495, \n",
    "    0.469158878505,\n",
    "    0.529962546816]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k3_lr3e = [\n",
    "    0.927102805521, \n",
    "    0.928971963508,\n",
    "    0.893258427413]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k5 = [\n",
    "    0.866459627329, \n",
    "    0.931464174455, \n",
    "    0.956386292835, \n",
    "    0.903125, \n",
    "    0.86875]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k7 = [\n",
    "    0.926086956522,\n",
    "    0.934782608955,\n",
    "    0.921739130435,\n",
    "    0.960869565736,\n",
    "    0.899122807018,\n",
    "    0.899122807018,\n",
    "    0.912280701754,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "k10 = [\n",
    "    0.938271604938,\n",
    "    0.95652173913,\n",
    "    0.913043478261,\n",
    "    0.91875,\n",
    "    0.94375,\n",
    "    0.9375,\n",
    "    0.90625,\n",
    "    0.925,\n",
    "    0.875,\n",
    "    0.9125]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sgd_k10_3e_15p = [\n",
    "    0.91975308642,\n",
    "    0.925465838509,\n",
    "    0.931677018634,\n",
    "    0.90625,\n",
    "    0.9625,\n",
    "    0.94375,\n",
    "    0.89375,\n",
    "    0.93125,\n",
    "    0.8375,\n",
    "    0.925]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sgd_k10_3e_20p = [\n",
    "    0.901234567901,\n",
    "    0.950310559006,\n",
    "    0.913043478261,\n",
    "    0.9125,\n",
    "    0.95625,\n",
    "    0.9375,\n",
    "    0.9,\n",
    "    0.93125,\n",
    "    0.8375,\n",
    "    0.93125]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Mean & Variance\n",
    "### k=3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9214346564676666\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(k3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.00018627990597391204\n"
     ]
    }
   ],
   "source": [
    "print (np.var(k3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3, lr = 1e-2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5099875156053333\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(k3_lr2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3, lr=3e-4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.916444398814\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(k3_lr3e))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### k=5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9052370189238002\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(k5))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0012287122363354827\n"
     ]
    }
   ],
   "source": [
    "print (np.var(k5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### k=7"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9220006539197143\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(k7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0004046008816595857\n"
     ]
    }
   ],
   "source": [
    "print (np.var(k7))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### k=10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9226586822328999\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(k10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.00048125837586399214\n"
     ]
    }
   ],
   "source": [
    "print (np.var(k10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3e-4, 15p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9176895943563\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(sgd_k10_3e_15p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3e-4, 15p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9170838605168001\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(sgd_k10_3e_20p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Adam"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k3_without_amsgrad = [\n",
    "    0.530841121495,\n",
    "    0.530841121495, \n",
    "    0.529962546816]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k3 = [\n",
    "    0.934579441035,\n",
    "    0.934579440144, \n",
    "    0.898876406727]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10 = [\n",
    "    0.901234567901,\n",
    "    0.937888198758,\n",
    "    0.925465838509,\n",
    "    0.925,\n",
    "    0.9625,\n",
    "    0.94375,\n",
    "    0.89375,\n",
    "    0.93125,\n",
    "    0.8625,\n",
    "    0.93125]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10_1e = [\n",
    "    0.913580246914,\n",
    "    0.925465838509,\n",
    "    0.937888198758,\n",
    "    0.8875,\n",
    "    0.9625,\n",
    "    0.93125,\n",
    "    0.8875,\n",
    "    0.91875,\n",
    "    0.84375,\n",
    "    0.93125]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10_0001 = [\n",
    "    0.925925925926,\n",
    "    0.937888198758,\n",
    "    0.919254658385,\n",
    "    0.9125,\n",
    "    0.94375,\n",
    "    0.94375,\n",
    "    0.925,\n",
    "    0.94375,\n",
    "    0.86875,\n",
    "    0.91875]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k3_0001_15p = [\n",
    "    0.925925925926,\n",
    "    0.950310559006,\n",
    "    0.913043478261,\n",
    "    0.93125,\n",
    "    0.95,\n",
    "    0.9625,\n",
    "    0.9,\n",
    "    0.9625,\n",
    "    0.85625,\n",
    "    0.9375]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k3_0001_15p_3rd = [\n",
    "    0.944444444444,\n",
    "    0.95652173913,\n",
    "    0.931677018634,\n",
    "    0.925,\n",
    "    0.9625,\n",
    "    0.95,\n",
    "    0.9,\n",
    "    0.925,\n",
    "    0.86875,\n",
    "    0.925]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10_0001_20p = [\n",
    "    0.91975308642,\n",
    "    0.95652173913,\n",
    "    0.931677018634,\n",
    "    0.925,\n",
    "    0.95,\n",
    "    0.95,\n",
    "    0.91875,\n",
    "    0.9375,\n",
    "    0.86875,\n",
    "    0.9125]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10_0001_25p = [\n",
    "    0.925925925926,\n",
    "    0.937888198758,\n",
    "    0.944099378882,\n",
    "    0.93125,\n",
    "    0.95,\n",
    "    0.94375,\n",
    "    0.89375,\n",
    "    0.95,\n",
    "    0.84375,\n",
    "    0.93125]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10_0001_30p = [\n",
    "    0.932098765432,\n",
    "    0.95652173913,\n",
    "    0.919254658385,\n",
    "    0.9125,\n",
    "    0.9625,\n",
    "    0.94375,\n",
    "    0.9125,\n",
    "    0.95625,\n",
    "    0.825,\n",
    "    0.91875]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k3_0001_10p = [\n",
    "    0.921495328885,\n",
    "    0.928971964399,\n",
    "    0.898876406057]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k3_0001_20p = [\n",
    "    0.921495328885,\n",
    "    0.934579440144,\n",
    "    0.889513109061]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k5_0001_10p = [\n",
    "    0.931677018634,\n",
    "    0.922118380062,\n",
    "    0.940809968847,\n",
    "    0.9125,\n",
    "    0.884375]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k7_0001_10p = [\n",
    "    0.939130434783,\n",
    "    0.930434783645,\n",
    "    0.939130434783,\n",
    "    0.960869565736,\n",
    "    0.855263157895,\n",
    "    0.899122807018,\n",
    "    0.868421052632]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k7_0001_20p = [\n",
    "    0.94347826087,\n",
    "    0.934782608696,\n",
    "    0.930434783645,\n",
    "    0.965217391823,\n",
    "    0.921052631579,\n",
    "    0.907894736842,\n",
    "    0.890350877193]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adam_k10_00001_10p = [\n",
    "    0.913580246914,\n",
    "    0.913043478261,\n",
    "    0.944099378882,\n",
    "    0.91875,\n",
    "    0.96875,\n",
    "    0.93125,\n",
    "    0.9,\n",
    "    0.906255,\n",
    "    0.85,\n",
    "    0.91875]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3 without amsgrad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5305482632686666\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k3_without_amsgrad))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3\n",
    "\n",
    "Train Log Loss Validation=  0.138096021399\n",
    "\n",
    "Test Log Loss Validation=  0.20165806206"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9226784293020001\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.01\n",
    "\n",
    "Train Log Loss Validation=  0.14229989114\n",
    "\n",
    "Test Log Loss Validation=  0.192718129101"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9214588605168\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=1e-3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9139434284181001\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10_1e))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=10\n",
    "Train Log Loss Validation=  0.117314350908\n",
    "\n",
    "Test Log Loss Validation=  0.181701044989"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9239318783068999\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10_0001))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=15\n",
    "Train Log Loss Validation=  0.123066677223\n",
    "\n",
    "Test Log Loss Validation=  0.178947531537"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9289279963193\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k3_0001_15p))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0009614523597267493\n"
     ]
    }
   ],
   "source": [
    "print (np.var(Adam_k3_0001_15p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=15\n",
    "Train Log Loss Validation=  0.107999644928\n",
    "\n",
    "Test Log Loss Validation=  0.174257478678"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9288893202208002\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k3_0001_15p_3rd))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=20\n",
    "Train Log Loss Validation=  0.10150099855\n",
    "\n",
    "Test Log Loss Validation=  0.179240771208"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9270451844183999\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10_0001_20p))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.0005789529924681881\n"
     ]
    }
   ],
   "source": [
    "print (np.var(Adam_k10_0001_20p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=25\n",
    "Train Log Loss Validation=  0.11924128702\n",
    "\n",
    "Test Log Loss Validation=  0.183048155108"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9251663503566\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10_0001_25p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=30\n",
    "Train Log Loss Validation=  0.112609400652\n",
    "\n",
    "Test Log Loss Validation=  0.179382493428"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9239125162947\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10_0001_30p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3, lr=0.0001, patience=10\n",
    "Train Log Loss Validation=  0.13753373181\n",
    "\n",
    "Test Log Loss Validation=  0.208624429906"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9164478997803333\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k3_0001_10p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3, lr=0.0001, patience=20\n",
    "Train Log Loss Validation=  0.129660983034\n",
    " \n",
    "Test Log Loss Validation=  0.207979759447"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9151959593633334\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k3_0001_20p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=5, lr=0.0001, patience=10\n",
    "Test Log Loss Validation=  0.196969180077"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9182960735086001\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k5_0001_10p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=7, lr=0.0001, patience=10\n",
    "Train Log Loss Validation=  0.13456575273\n",
    "\n",
    "Test Log Loss Validation=  0.204074223575"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9131960337845714\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k7_0001_10p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=7, lr=0.0001, patience=20\n",
    "Train Log Loss Validation=  0.113610347708\n",
    "\n",
    "Test Log Loss Validation=  0.193107086096"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9276016129497142\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k7_0001_20p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10, lr=0.0001, patience=10\n",
    "\n",
    "Train Log Loss Validation=  0.124790371608\n",
    "\n",
    "Test Log Loss Validation=  0.213980056827"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9164478104056999\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adam_k10_00001_10p))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Adadelta"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adadelta_k10 = [\n",
    "    0.901234567901,\n",
    "    0.888198757764,\n",
    "    0.900621118012,\n",
    "    0.91875,\n",
    "    0.975,\n",
    "    0.91875,\n",
    "    0.85625,\n",
    "    0.91875,\n",
    "    0.85,\n",
    "    0.90625]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10\n",
    "Train Log Loss Validation=  0.193596459305\n",
    "\n",
    "Test Log Loss Validation=  0.236750899909"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.9033804443676999\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adadelta_k10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Adamax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adamax_k3 = [\n",
    "    0.530841121495,\n",
    "    0.904494384255,\n",
    "    0.927102805967]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "Adamax_k10 = [\n",
    "    0.530864197531,\n",
    "    0.937888198758,\n",
    "    0.937888198758,\n",
    "    0.9375,\n",
    "    0.53125,\n",
    "    0.9375,\n",
    "    0.9125,\n",
    "    0.9375,\n",
    "    0.53125,\n",
    "    0.925]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7874794372389999\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adamax_k3))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## k=10\n",
    "\n",
    "Train Log Loss Validation=  0.271070528743\n",
    "\n",
    "Test Log Loss Validation=  3.42418115748"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.8119140595047\n"
     ]
    }
   ],
   "source": [
    "print (np.mean(Adamax_k10))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Compare"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEyCAYAAAALJfw2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1125f37f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3, 4]\n",
    "y = [0.811914059505, 0.903380444368, 0.921458860517, 0.922658682233]\n",
    "labels = ['Adamax', 'Adadelta', 'Adam', 'SGD']\n",
    "\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAEyCAYAAAD5tWvWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1125f3710>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3, 4]\n",
    "y = [3.42418115748, 0.236750899909, 0.192718129101, 0.193213699335]\n",
    "labels = ['Adamax', 'Adadelta', 'Adam', 'SGD']\n",
    "\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAEiCAYAAADEasRGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/l/AzwwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1125f35f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2]\n",
    "y = [0.921458860517, 0.922658682233]\n",
    "labels = ['Adam', 'SGD']\n",
    "\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEyCAYAAAALJfw2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ni4iIqJSwiIiISgmLiIiolLCIiIhK/x9ApkXq7qLGtwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112891898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3]\n",
    "y = [ 0.236750899909, 0.192718129101, 0.193213699335]\n",
    "labels = ['Adadelta', 'Adam', 'SGD']\n",
    "\n",
    "plt.title('log loss validation')\n",
    "plt.ylabel('loss validation')\n",
    "plt.xlabel('optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAEiCAYAAADEasRGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIhqRBBMREY1IgomIiEb8D9RAJ7fkLv0eAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1125dce48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2]\n",
    "y = [0.192718129101, 0.193213699335]\n",
    "labels = ['Adam', 'SGD']\n",
    "\n",
    "plt.title('log loss validation')\n",
    "plt.ylabel('log loss score')\n",
    "plt.xlabel('optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Learning rate = 1e-3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAEyCAYAAAD5tWvWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1125f34a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3, 4]\n",
    "y = [3.42418115748, 0.236750899909, 0.211348751952, 0.193213699335]\n",
    "labels = ['Adamax', 'Adadelta', 'Adam', 'SGD']\n",
    "\n",
    "plt.title('Log loss validation')\n",
    "plt.ylabel('Log loss score')\n",
    "plt.xlabel('Optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEyCAYAAAALJfw2AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAHphJREFUeJzt3XucHXWd5vHPk0CEyH2IjiYkAURm\ngoOgTfCOYkTQIagggo0IMmZnJOMqs+6A6Iq4rArecEClnQFHN4oE1EXk5otBRx1ROoBo0EgIJISL\nBBCJRi6BZ/+oajg03V0nna6u0yfP+/Xq1zn1q8v5di79dP1+VfWTbSIiIkYyqekCIiKi8yUsIiKi\nUsIiIiIqJSwiIqJSwiIiIiolLCIiolLCIjYpkk6R9H+brqOKpK9I+t/l+1dKWtbOtqP8rD9K2mW0\n+8emIWERHUPSbZLmNV1Hp7H9I9u7j8WxJP1A0t8NOv5WtleMxfGjeyUsIiKiUsIiJgRJ75a0XNL9\nki6W9NyWdQdIWibpD5K+IOmHg397HuG48yUtlfRA+Vv3X7es+2dJd0haWx7/tWX7XEn9kh6U9DtJ\nnxnm2L+W9Lcty5tJulfSi8rlxZLuLuv+T0l7DHOcV0ta3bK8t6Tryrq+CWzRsm57SZdIWiPp9+X7\nGeW604BXAmeVXU9nle2W9Lzy/baSvlruv1LShyRNKtcdI+nHkj5VHvtWSQe18+ccE1/CIjqepP2B\njwOHA88BVgLnl+t2BC4ETgL+AlgGvKzN4z4f+AbwPmAacCnwXUlTJO0OLAT2sb018HrgtnLXM4Ez\nbW8D7ApcMMxHfAM4smX59cC9tq8rly8DdgOeBVwHLGqj5inAd4CvATsAi4FDWzaZBJwHzAJmAn8G\nzgKwfTLwI2Bh2fW0cIiP+BdgW2AXYD/gaODYlvX7UvwZ7wicDvybJFXVHRNfwiImgl7gXNvX2X6Y\nIhheKmk28AZgqe1v2V4PfB64u83jvg34nu3v234U+BSwJUXYPAY8A5gjaXPbt9m+pdzvUeB5kna0\n/Ufb1wxz/K8D8yVNLZffXrYBYPtc22vL7+kU4IWStq2o+SXA5sDnbD9q+0Lg2pZj3mf7ItvrbK8F\nTqP4oV9J0uTyz+Sksq7bgE8D72jZbKXtL9t+DPh3ivB+djvHj4ktYRETwXMpziYAsP1H4D5gernu\n9pZ1BlYPPkCbx328PNZ028spzjhOAe6RdH5L19dxwPOB30i6trWrqVV5jF8DB5eBMZ8yLCRNlvQJ\nSbdIepAnz1p2bKPmO/zUJ4A+8T1ImirpnLIL6UHgP4HtyiCosiMwpfV45fvpLctPBLHtdeXbrdo4\ndkxwCYuYCO6k6FYBQNIzKbqc7gDuAma0rFPr8gYeV8BO5XGx/XXbryi3MfDJsv1m20dSdB99Eriw\nrGkoA11RhwA3lQECxVnGIcA8im6f2QNlVNR8FzB9UNfPzJb3/wTsDuxbdpO9atBxR3rM9L0UZ02z\nWtpmUv55xKYtYRGdZnNJW7R8bUbx2/ixkvaS9Azg/wA/K7tJvgf8jaQ3ldseD/xlm591AfBGSa+V\ntDnFD9qHgf+StLuk/cvPe4ii7/8xAElHSZpWnok8UB7rsWE+43zgAOAfaOmCArYuP+s+YGr5PbXj\np8B64L3lgPlbgLmDjvtn4AFJOwAfGbT/7yjGI56m7Fq6ADhN0taSZgEnAB1/X0rUL2ERneZSih92\nA1+n2L4K+DBwEcVv1rsCRwDYvhd4K8Vg633AHKCf4gfxiGwvA46iGNS9FzgYONj2IxTjFZ8o2++m\nOIv4YLnrgcBSSX+kGOw+wvZDw3zGXRQ/4F8GfLNl1VcpunjuAG4Chhv3GHy8R4C3AMcAv6cYY/hW\nyyafoxh3ubc85uWDDnEmcFh5NdPnh/iIfwT+BKwAfkwRcOe2U1t0N2Xyo+gm5WWeq4Fe21c3XU9E\nt8iZRUx4kl4vabuyy+iDFP3zbf2mHhHtSVhEN3gpcAtPdiW9yfafmy0porukGyoiIirlzCIiIiol\nLCIiotJmTRcwVnbccUfPnj276TIiIiaUJUuW3Gt7WtV2XRMWs2fPpr+/v+kyIiImFEkrq7dKN1RE\nRLQhYREREZUSFhERUSlhERERlRIWERFRKWERERGVEhYREVEpYREREZUSFhERUSlhERERlRIWERFR\nKWERERGVEhYREVGp1rCQdKCkZZKWSzpxiPUnSLpJ0o2SrpI0a9D6bSTdIemsOuuMiIiR1RYWkiYD\nZwMHAXOAIyXNGbTZ9UCP7T2BC4HTB63/GPDDumqMiIj21HlmMRdYbnuF7UeA84FDWjewfbXtdeXi\nNcCMgXWSXgw8G7iyxhojIqINdYbFdOD2luXVZdtwjgMuA5A0Cfg08IHaqouIiLbVOVOehmjzkBtK\nRwE9wH5l03uAS23fLg11mCf2WwAsAJg5c+ZGFRsREcOrMyxWAzu1LM8A7hy8kaR5wMnAfrYfLptf\nCrxS0nuArYApkv5o+ymD5Lb7gD6Anp6eIYMoIiI2Xp1hcS2wm6SdgTuAI4C3t24gaW/gHOBA2/cM\ntNvubdnmGIpB8KddTRUREeOjtjEL2+uBhcAVwK+BC2wvlXSqpPnlZmdQnDkslnSDpIvrqiciIkZP\ndnf03vT09Li/v7/pMiIiJhRJS2z3VG2XO7gjIqJSwiIiIiolLCIiolLCommLFsHs2TBpUvG6aFHT\nFUVEPE2dl85GlUWLYMECWFc+8WTlymIZoLd3+P0iIsZZziyadPLJTwbFgHXrivaIiA6SsGjSqlUb\n1h4R0ZCERZOGe55VnnMVER0mYdGk006DqVOf2jZ1atEeEdFBEhZN6u2Fvj6YNQuk4rWvL4PbEdFx\ncjVU03p7Ew4R0fFyZhEREZUSFhERUSlhERERlRIWERFRKWERERGVEhYREVEpYREREZUSFhERUSlh\nERERlRIWERFRKWERERGVEhYREVEpYREREZUSFhERUSlhERERlWoNC0kHSlomabmkE4dYf4KkmyTd\nKOkqSbPK9lmSlki6QdJSSX9fZ50RETGy2sJC0mTgbOAgYA5wpKQ5gza7HuixvSdwIXB62X4X8DLb\newH7AidKem5dtUZExMjqPLOYCyy3vcL2I8D5wCGtG9i+2va6cvEaYEbZ/ojth8v2Z9RcZ0REVKjz\nh/B04PaW5dVl23COAy4bWJC0k6Qby2N80vadg3eQtEBSv6T+NWvWjFHZERExWJ1hoSHaPOSG0lFA\nD3DGExvat5fdU88D3inp2U87mN1nu8d2z7Rp08ao7IiIGKzOsFgN7NSyPAMY6uxgHnAyML+l6+kJ\n5RnFUuCVNdUZEREV6gyLa4HdJO0saQpwBHBx6waS9gbOoQiKe1raZ0jasny/PfByYFmNtUZExAg2\nq+vAttdLWghcAUwGzrW9VNKpQL/tiym6nbYCFksCWGV7PvDXwKclmaI761O2f1lXrRERMTLZQw4j\nTDg9PT3u7+9vuoyIiAlF0hLbPVXb5ZLUiIiolLCIiIhKCYuIiKiUsIiIiEoJi4iIqJSwiIiISgmL\niIiolLCIiIhKCYuIiKiUsIiIiEoJi4iIqJSwiIiISgmLiIioVBkWkp4v6SpJvyqX95T0ofpLi4iI\nTtHOmcWXgZOARwFs30gxkVFERGwi2gmLqbZ/PqhtfR3FREREZ2onLO6VtCtgAEmHAXfVWlVERHSU\ndqZVPR7oA/5K0h3ArUBvrVVFRERHGTEsJE0CemzPk/RMYJLtteNTWkREdIoRu6FsPw4sLN//KUER\nEbFpamfM4vuS/oeknSTtMPBVe2UREdEx2hmzeFf5enxLm4Fdxr6ciIjoRJVhYXvn8SgkIiI6V2VY\nSNoc+AfgVWXTD4BzbD9aY10REdFB2umG+iKwOfCFcvkdZdvf1VVURER0lnYGuPex/U7b/1F+HQvs\n087BJR0oaZmk5ZJOHGL9CZJuknRj+fypWWX7XpJ+Kmlpue5tG/ZtRUTEWGonLB4r7+AGQNIuwGNV\nO0maDJwNHATMAY6UNGfQZtdT3MexJ3AhcHrZvg442vYewIHA5yRt10atERFRg3a6oT4AXC1pBSBg\nFnBsG/vNBZbbXgEg6XzgEOCmgQ1sX92y/TXAUWX7b1u2uVPSPcA04IE2PjciIsZYO1dDXSVpN2B3\nirD4je2H2zj2dOD2luXVwL4jbH8ccNngRklzgSnALW18ZkRE1KCd+SyOB7a0faPtXwBTJb2njWNr\niDYP8xlHAT3AGYPanwN8DTi2vJt88H4LJPVL6l+zZk0bJUVExGi0M2bxbttPdP/Y/j3w7jb2Ww3s\n1LI8A7hz8EaS5gEnA/Nbz1gkbQN8D/iQ7WuG+gDbfbZ7bPdMmzatjZIiImI02gmLSZKeOEsoB66n\ntLHftcBuknaWNIViwqSLWzeQtDdwDkVQ3NPSPgX4NvBV24vb+KyIiKhRO2FxBXCBpNdK2h/4BnB5\n1U6211M8hPAK4NfABbaXSjpV0vxyszOArYDFkm6QNBAmh1PcBHhM2X6DpL027FuLiIixInvIYYQn\nNygeU74AmEcxDnEl8K+2Ky+fHU89PT3u7+9vuoyIiAlF0hLbPVXbtXM11OPAl4AvlU+bndFpQRER\nEfVq52qoH0japgyKG4DzJH2m/tIiIqJTtDNmsa3tB4G3AOfZfjFFl1RERGwi2gmLzcr7HQ4HLqm5\nnoiI6EDthMWpFFc0Lbd9bflsqJvrLSsiIjpJOwPci4HFLcsrgEPrLCoiIjpLO2cWERGxiUtYRERE\npYRFRERUauc+i/9e3mchSf8m6TpJB4xHcRER0RnaObN4V3mfxQEUExAdC3yi1qoiIqKjtBMWA0+c\nfQPFTXm/YOi5KiIioku1ExZLJF1JERZXSNoaeNpERBER0b3amYP7OGAvYIXtdeUzotqZgzsiIrpE\nO2cWLwWW2X6gnP70Q8Af6i0rIiI6STth8UVgnaQXAv8TWAl8tdaqIiKio7QTFutdzJB0CHCm7TOB\nrestKyIiOkk7YxZrJZ0EvAN4ZTkH9+b1lhUREZ2knTOLtwEPU9xvcTcwnWLu7IiI2ERUhkUZEIuA\nbSX9LfCQ7YxZRERsQtp53MfhwM+Bt1JMgPQzSYfVXVhERHSOdrqhTgb2sf1O20cDc4EP11tWRJdY\ntAhmz4ZJk4rXRYuarihiVNoZ4J5k+56W5fvI02ojqi1aBAsWwLp1xfLKlcUyQG9vc3VFjEI7P/Qv\nl3SFpGMkHQN8D7i03rIiusDJJz8ZFAPWrSvaIyaYdqZV/YCkQ4GXUzxAsM/2t2uvLGKiW7Vqw9oj\nOlg73VDYvgi4qOZaIrrLzJlF19NQ7RETzLDdUJLWSnpwiK+1kh5s5+CSDpS0TNJySScOsf4ESTdJ\nulHSVZJmtay7XNIDki4Z3bcW0bDTToOpU5/aNnVq0R4xwQwbFra3tr3NEF9b296m6sDlnd5nAwcB\nc4AjJc0ZtNn1QI/tPYELgdNb1p1Bcdd4xMTU2wt9fTBrFkjFa19fBrdjQqrzqqa5wHLbK2w/ApxP\n8XypJ9i+2vbACOA1wIyWdVcBa2usL6J+vb1w223w+OPFa4IiJqg6w2I6cHvL8uqybTjHAZdtyAdI\nWiCpX1L/mjVrRlFiRES0o86wGGrqVQ+5YTFPRg8b+Mwp2322e2z3TJs2bRQlRkREO9q6GmqUVgM7\ntSzPAO4cvJGkeRR3ie9n++Ea64mIiFFq59lQQ10Vdbukb0vaZYRdrwV2k7SzpCnAEcDFg469N3AO\nMH/QXeIREdFB2jmz+AzFGcHXKbqWjgD+ElgGnAu8eqidbK+XtBC4ApgMnGt7qaRTgX7bF1N0O20F\nLJYEsMr2fABJPwL+CthK0mrgONtXjPYbjYiI0VMxCd4IG0g/s73voLZrbL9E0i9sv7DWCtvU09Pj\n/v7+psuIiJhQJC2x3VO1XTsD3I9LOlzSpPLr8JZ1IydNRER0hXbCopfi5rh7yq93AEdJ2hJYWGNt\nERHRIdp5kOAK4OBhVv94bMuJiIhO1M7VUDPKK5/ukfQ7SRdJmlG1X0REdI92uqHOo7jk9bkUd2B/\nt2yLiIhNRDthMc32ebbXl19fAXK7dETEJqSdsLhX0lGSJpdfR1FMrRoREZuIdsLiXcDhwN3AXcBh\nwLF1FhUREZ2lMixsr7I93/Y028+y/SbgLeNQW0REdIjRPnX2hDGtIiIiOtpow2Kox49HRESXGm1Y\n5DEfERGbkGHv4Ja0lqFDQcCWtVUUEREdZ9iwsL31eBYSERGdq85pVSMiokskLCIiolLCIiIiKiUs\nIiKiUsIiIiIqJSwiIqJSwiIiIiolLCIiolLCIiIiKiUsIiKiUsIiIiIq1RoWkg6UtEzSckknDrH+\nBEk3SbpR0lWSZrWse6ekm8uvd9ZZZ0REjKy2sJA0GTgbOAiYAxwpac6gza4HemzvCVwInF7uuwPw\nEWBfYC7wEUnb11VrRESMrM4zi7nActsrbD8CnA8c0rqB7attrysXrwFmlO9fD3zf9v22fw98Hziw\nxlojImIEdYbFdOD2luXVZdtwjgMuG+W+ERFRo2HnsxgDQ029OuQMe5KOAnqA/TZkX0kLgAUAM2fO\nHF2VERFRqc4zi9XATi3LM4A7B28kaR5wMjDf9sMbsq/tPts9tnumTZs2ZoVHRMRT1RkW1wK7SdpZ\n0hTgCODi1g0k7Q2cQxEU97SsugI4QNL25cD2AWVbREQ0oLZuKNvrJS2k+CE/GTjX9lJJpwL9ti8G\nzgC2AhZLAlhle77t+yV9jCJwAE61fX9dtUZExMhkDzmMMOH09PS4v7+/6TIiIiYUSUts91Rtlzu4\nIyKiUsIiIiIqJSwiIqJSwiIiIiolLCIiolLCIiIiKiUsIiKiUsIiIiIqJSwiIqJSwiIiIiolLCIi\nolLCIiIiKiUsIiKiUsIiIiIqJSwiIqJSwiIiYjQWLYLZs2HSpOJ10aKmK6pVbTPlRUR0rUWLYMEC\nWLeuWF65slgG6O1trq4a5cwiImJDnXzyk0ExYN26or1LJSwiIjbUqlUb1t4FEhYRERtq5swNa+8C\nCYuIiA112mkwdepT26ZOLdq7VMIiImJD9fZCXx/MmgVS8drX17WD25CroSIiRqe3t6vDYbCcWURE\nRKWERUREVKo1LCQdKGmZpOWSThxi/askXSdpvaTDBq37pKRflV9vq7POiIgYWW1hIWkycDZwEDAH\nOFLSnEGbrQKOAb4+aN83Ai8C9gL2BT4gaZu6ao2IiJHVeWYxF1hue4XtR4DzgUNaN7B9m+0bgccH\n7TsH+KHt9bb/BPwCOLDGWiMiYgR1hsV04PaW5dVlWzt+ARwkaaqkHYHXADuNcX0REdGmOi+d1RBt\nbmdH21dK2gf4L2AN8FNg/dM+QFoALACY2cV3TkZENK3OM4vVPPVsYAZwZ7s72z7N9l62X0cRPDcP\nsU2f7R7bPdOmTdvogiMiYmh1hsW1wG6SdpY0BTgCuLidHSVNlvQX5fs9gT2BK2urNCIiRlRbN5Tt\n9ZIWAlcAk4FzbS+VdCrQb/visqvp28D2wMGSPmp7D2Bz4EeSAB4EjrL9tG6oiIgYH7U+7sP2pcCl\ng9r+V8v7aym6pwbv9xDFFVEREdEBcgd3RERUSlhERESlhEVERFRKWERERKWERUREVEpYREREpYRF\nRERUSlhERESlhEVERFRKWERERKWERUREVEpYREREpYRFRERUSlhEREQl2W3NdNrxJK0BVjZdx0bY\nEbi36SKiNvn77V4T/e92lu3KqUa7JiwmOkn9tnuariPqkb/f7rWp/N2mGyoiIiolLCIiolLConP0\nNV1A1Cp/v91rk/i7zZhFRERUyplFRERUSlhERESlhEVERFTarOkCIrqVpO2BnWj5f2b7uuYqirEk\naRqA7TVN1zIeEhYNkbQb8HFgDrDFQLvtXRorKsaMpI8BxwC3AANXkRjYv6maYuNJEvARYCEgYJKk\n9cC/2D610eJqlrBoznkU/+g+C7wGOJbiH190h8OBXW0/0nQhMabeB7wc2Mf2rQCSdgG+KOn9tj/b\naHU1yphFc7a0fRXF5csrbZ9CfuvsJr8Ctmu6iBhzRwNHDgQFgO0VwFHluq6VM4vmPCRpEnCzpIXA\nHcCzGq4pxs7Hgesl/Qp4eKDR9vzmSooxsLntpz000PYaSZs3UdB4SVg0533AVOC9wMcouqK6+jeT\nTcy/A58Efgk83nAtMXZG6lbs6i7H3MHdEElvtb24qi0mJkk/tL1f03XE2JL0GPCnoVYBW9ju2rOL\nhEVDJF1n+0VVbTExSfoMRffTxTy1GyqXzsaElG6ocSbpIOANwHRJn29ZtQ2wvpmqogZ7l68vaWnL\npbNdRtJ0YHK5eKftrv0/nLAYf3cCS4D55euAtcD7G6koxpzt1zRdQ4w9SSdRDHIP3FPxU+APwOYU\n41Qfb6q2uqUbqiGSNuvm30ICJL0R2IOn3nTZ1TdudTtJ1wGvtP2ncvl623tLmgz80PYrmq2wPjmz\nGGeSfkl5R29xM+hT2d5zvGuKsSfpSxRXu70G+FfgMODnjRYVY2IgKEpnlm2PSdqyoZLGRc4sxpmk\nWSOtt71yvGqJ+ki60faeLa9bAd+yfUDTtcXoSfotsIftRwe1PwP4le3dmqmsfrmDe5yVd2uvbAmF\n3cr39wD3N1hajK0/l6/rJD0XeBTYucF6YmxcCJwjaepAg6RnAueU67pWwqIhkt5N+Q+vbJoBfKe5\nimKMXSJpO+AM4DrgNuD8RiuKsfBhil/sVklaImkJxd/t3eW6rpVuqIZIugGYC/zM9t5l2y9t/02z\nlcVYK7sotrD9h6ZriY0jaR9gNfAA8Dzg1cDBwG+AU2x3be9ABrib87DtRwYGuSVtxpOPso4JStJb\nRliH7W+NZz0x5s4B5tn+czlfyUnAPwJ7AX0UFzJ0pYRFc34o6YPAlpJeB7wH+G7DNcXGO7h8fRbw\nMuA/yuXXAD8AEhYT2+SWs4e3AX22LwIuKnsLulbGLJpzIrCG4kFz/w24FPhQoxXFRrN9rO1jKc4S\n59g+1PahFPdbxMQ3uewFAHgtT/4yAF3+y3dXf3OdzPbjwJfLr+g+s23f1bL8O2D3poqJMfMNil6B\neymuePsRgKTnUdzJ3bUywD3OWm/KG0puyusOks4CdqP44WLgCOBm2+9ttLDYaJJeAjwHuLLlTu7n\nA1t184MiExbjrOWmvOPL16+Vr73AujwOontIejPwqnLx98CzbR8/wi4RHSth0RBJP7H98qq2mLgk\n7QW8nWI+7luBi2yf1WxVEaOTMYvmPFPSK2z/GEDSy4BnNlxTbKSyO+II4EjgPuCbFL+U5Sm0MaHl\nzKIhkl4MnAtsWzY9ALyrm/s8NwWSHqcY9DzO9vKybYXtXZqtLGLj5MyiIbaXAC+UtA1FaHf1lRSb\nkEMpziyulnQ5xSM+nv544YgJJmcWDcp8B92rfLjcmyi6o/anmBjn27avbLSwiFFKWDRkuPkObB/X\naGEx5iTtALwVeJvtTKsaE1LCoiGZ7yAiJpI87qM5me8gIiaMDHA3Z/B8B6bojoqI6DjphuoAme8g\nIjpdwmKcjTTfAZD5DiKiI6UbavxlvoOImHASFuOsnOsASZdQzHdwV7n8HODsJmuLiBhOroZqTuY7\niIgJI2cWzfmBpCt46nwHVzVbUkTE0DLA3aDMdxARE0W6oZp1K8XNeG+mGOD+dbPlREQMLd1Q4yzz\nHUTERJRuqHGW+Q4iYiJKN9T4OxS4m2K+gy9Lei2Z7yAiOlzOLBqS+Q4iYiJJWHSAzHcQEZ0uYRER\nEZUyZhEREZUSFhERUSlhEVGSNEPS/5N0s6RbJJ0pacoI228n6T0ty8+VdOEGfuapkuZtTN0R4yFj\nFhGAJAE/A75o+zxJk4E+4H7bHxhmn9nAJbZfMG6FDl3HZrbXN1lDdL/cwR1R2B94yPZ5ALYfk/R+\n4FZJtwKvB55BMU/6121/FPgEsKukG4DvUzxi/hLbL5B0DMWl0ZOBFwCfBqYA7wAeBt5g+35JXwEu\nAW7jyWl1JwMvsC1Ju5bHnQasA95t+zflfvcDe1NMy/tPdf3BREDCImLAHsCS1gbbD0paRfH/ZC7F\nD/11wLWSvgecSPFDfS944kyj1QsofphvASwH/tn23pI+CxwNfK7ls/qBgeOcAVxeruoD/t72zZL2\nBb5AEWwAzwfm2X5so7/7iAoJi4iCKB4VP1z7923fByDpW8ArgO9UHPNq22uBtZL+AHy3bP8lsOeQ\nRUiHAy8CDpC0FcVsiouLXjKgOLsZsDhBEeMlYRFRWErxKJYnSNoG2Al4jKcHSTuDfQ+3vH+8Zflx\nhvi/J2kP4KPAq8pusEnAAwNnLkP4Uxs1RIyJXA0VUbgKmCrpaIBygPvTwFcoup5eJ2kHSVtSjEX8\nBFgLbD0WHy5pW+B84Gjba6DoBqMYM3lruY0kvXAsPi9iQyUsIgAXlwW+GXirpJuB3wIPAR8sN/kx\n8DXgBuAi2/1lt9RPJP2qHGfYGG8CZgFflnRDOWgO0AscJ+kXFGc/h2zk50SMSi6djahQXtnUY3th\n07VENCVnFhERUSlnFhERUSlnFhERUSlhERERlRIWERFRKWERERGVEhYREVEpYREREZX+PzFzbMGq\nwH8jAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112a3f0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3]\n",
    "y = [0.236750899909, 0.211348751952, 0.193213699335]\n",
    "labels = ['Adadelta', 'Adam', 'SGD']\n",
    "\n",
    "plt.title('Log loss validation')\n",
    "plt.ylabel('Log loss score')\n",
    "plt.xlabel('Optimizer')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEuCAYAAAB/MT7WAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIiIiKv1/wmDA66HlrW0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x112995a20>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3, 4]\n",
    "y = [0.20165806206, 0.234735911185, 0.181701044989, 0.213980056827]\n",
    "labels = ['0.01', '0.001', '0.0001', '0.00001']\n",
    "\n",
    "plt.title('Log Loss Validation')\n",
    "plt.ylabel('Log Loss Score')\n",
    "plt.xlabel('Learning Rate')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAESCAYAAADXMlMiAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11297c0f0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x = [1, 2, 3, 4, 5]\n",
    "y = [0.181701044989, 0.174257478678, 0.179240771208, 0.183048155108, 0.179382493428]\n",
    "labels = ['10', '15', '20', '25', '30']\n",
    "\n",
    "plt.title('Log Loss Validation')\n",
    "plt.ylabel('Log Loss Score')\n",
    "plt.xlabel('Patience')\n",
    "\n",
    "plt.plot(x, y, 'ro')\n",
    "# You can specify a rotation for the tick labels in degrees or with keywords.\n",
    "plt.xticks(x, labels, rotation='vertical')\n",
    "# Pad margins so that markers don't get clipped by the axes\n",
    "plt.margins(0.2)\n",
    "# Tweak spacing to prevent clipping of tick-labels\n",
    "plt.subplots_adjust(bottom=0.15)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# x = np.arange(5)\n",
    "# plt.bar(x, height= [0.181701044989, 0.174257478678, 0.179240771208, 0.183048155108, 0.179382493428])\n",
    "# plt.xticks(x+.5, ['10','20','30']);\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}