Fashion-MNSIT.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FASHION MNIST USING LOGISTIC REGRESSION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "df=pd.read_csv('fashion-mnist_test.csv')\n",
    "dft=pd.read_csv('fashion-mnist_train.csv')\n",
    "vt=dft.values\n",
    "v=df.values\n",
    "X_test = vt[:,1:]\n",
    "y_test = vt[:,0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10000,)\n",
      "(10000, 784)\n"
     ]
    }
   ],
   "source": [
    "y=v[:,0]\n",
    "x=v[:,1:]\n",
    "print(y.shape)\n",
    "print(x.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21e15a2f2b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8\n"
     ]
    }
   ],
   "source": [
    "#Visualise SOme Samples\n",
    "\n",
    "def drawImg(sample):\n",
    "    img = sample.reshape((28,28))\n",
    "    plt.imshow(img,cmap='gray')\n",
    "    plt.show()\n",
    "    \n",
    "drawImg(x[14])\n",
    "print(y[14])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "model = LogisticRegression()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
       "          intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
       "          penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
       "          verbose=0, warm_start=False)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "model.fit(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.76871666666666671"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### score for one vs rest\n",
    "model.score(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6, 9, 6, ..., 8, 8, 7], dtype=int64)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([6, 9, 6, 2, 3], dtype=int64)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_test[:5,:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2, 9, 6, 0, 3], dtype=int64)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# FASHION MNIST USING KNN from scratch\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "\n",
    "def dist(x1,x2):\n",
    "    return np.sqrt(sum((x1-x2)**2))\n",
    "\n",
    "def knn(X,Y,queryPoint,k=5):\n",
    "    \n",
    "    vals = []\n",
    "    m = X.shape[0]\n",
    "    \n",
    "    for i in range(m):\n",
    "        d = dist(queryPoint,X[i])\n",
    "        vals.append((d,Y[i]))\n",
    "        \n",
    "    \n",
    "    vals = sorted(vals)\n",
    "    # Nearest/First K points\n",
    "    vals = vals[:k]\n",
    "    \n",
    "    vals = np.array(vals)\n",
    "    \n",
    "    #print(vals)\n",
    "    \n",
    "    new_vals = np.unique(vals[:,1],return_counts=True)\n",
    "    #print(new_vals)\n",
    "    \n",
    "    index = new_vals[1].argmax()\n",
    "    pred = new_vals[0][index]\n",
    "    \n",
    "    return pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9\n"
     ]
    }
   ],
   "source": [
    "pred = knn(x,y,X_test[1])\n",
    "\n",
    "print(int(pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x21e16c95438>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9\n"
     ]
    }
   ],
   "source": [
    "drawImg(X_test[1])\n",
    "print(y_test[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}