The Algorithms logo
The Algorithms
À proposFaire un don

Clothing Detection

H
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#Importing Packages\n",
    "from tensorflow import keras \n",
    "import numpy as np           \n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
      "32768/29515 [=================================] - 0s 3us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
      "26427392/26421880 [==============================] - 25s 1us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
      "8192/5148 [===============================================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
      "4423680/4422102 [==============================] - 4s 1us/step\n"
     ]
    }
   ],
   "source": [
    "#Importing Keras Dataset\n",
    "data = keras.datasets.fashion_mnist\n",
    "\n",
    "#train and test data segregation\n",
    "(train_images, train_labels), (test_images, test_labels) = data.load_data()\n",
    "\n",
    "class_names = [\"T-shirt/top\", \"Trouser\", \"Pullover\", \"Dress\", \"Coat\", \"Sandal\", \"Shirt\", \"Sneaker\", \"Bag\", \"Ankle Boot\"]\n",
    "\n",
    "train_images = train_images/255.0\n",
    "test_images = test_images/255.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From C:\\Users\\vinay\\Anaconda3\\lib\\site-packages\\tensorflow\\python\\ops\\init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Call initializer instance with the dtype argument instead of passing it to the constructor\n",
      "Epoch 1/5\n",
      "60000/60000 [==============================] - 8s 132us/sample - loss: 0.4980 - acc: 0.8253\n",
      "Epoch 2/5\n",
      "60000/60000 [==============================] - 8s 127us/sample - loss: 0.3717 - acc: 0.8662\n",
      "Epoch 3/5\n",
      "60000/60000 [==============================] - 6s 93us/sample - loss: 0.3341 - acc: 0.8787\n",
      "Epoch 4/5\n",
      "60000/60000 [==============================] - 8s 130us/sample - loss: 0.3126 - acc: 0.8852\n",
      "Epoch 5/5\n",
      "60000/60000 [==============================] - 6s 92us/sample - loss: 0.2928 - acc: 0.8929\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x25f99983fd0>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Create a object of model class\n",
    "model = keras.Sequential([\n",
    "    keras.layers.Flatten(input_shape=(28,28)),\n",
    "    keras.layers.Dense(128, activation=\"relu\"),\n",
    "    keras.layers.Dense(10, activation=\"softmax\")\n",
    "])\n",
    "#Compile the model with MSE loss and Adam optimizer\n",
    "model.compile(optimizer=\"adam\", loss=\"sparse_categorical_crossentropy\", metrics=[\"accuracy\"])\n",
    "#fitting the model\n",
    "model.fit(train_images, train_labels, epochs=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000/10000 [==============================] - 1s 79us/sample - loss: 0.3884 - acc: 0.8587\n",
      "Accuracy:  0.8587\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAEWCAYAAACg3+FOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAXnklEQVR4nO3df7RdZX3n8feHBAgkAfKLTAiBqxjKWHUCvc2aVayDlWERxgpZamu6pKGjYld1WdZokdVxNI7jwHJqXXasRQQqtkhLIRTKUhJgyAABY64IJCEVEG8gP8gvID8AgYTv/LGfOIfrOc++Oefcey48n9daZ91z9nP23t+z7/3cfc559rO3IgIze+M7pNcFmNnocNjNCuGwmxXCYTcrhMNuVgiH3awQDnsPSeqTFJLGp8c/kLS4jeWcIGmvpHHdr7J9Q1/fcNtsZDjsNSQNSnoxhWmrpL+VNGkk1hURCyLimmHWdGbDfE9GxKSI2D8SdanyhKRHRmL5B0vSEkmvpN/JXknrJb2/S8teIemj3VjWWOOwD8/vRsQk4DTgN4HPDX1CCsQbdXu+CzgWeLOk3+x1Mck/pn9wk4CLgL+XNLPXRY1lb9Q/zhEREZuAHwBvg1/uBb4saSXwAlUYjpZ0laQtkjZJ+h8H3l5LGifpLyTtkPQE8J8alz90ryLpY2mvtUfSI5JOk/R3wAnAv6S92sVNPg4cJ+kWSc9IelzSxxqWuUTS9ZK+m5a7TlJ/zUtfDNwMfD/dH1rzlyStTMtbLml6s4VIen96V/K2Jm0tt1udiFgG7AFOGrLtHk/b4BZJxzW0/Zak1ZJ2pZ+/laZ/Gfht4Btp235jOOt/3YgI3zI3YBA4M92fA6wDvpQerwCeBH4dGA8cCvwz8C1gItXe8EfAx9Pz/xj417ScqcBdQADjG5b30XT/g8AmqncSAt4CnDi0pvS4b8hy/i/wTWACMA/YDrwntS0BfgGcA4wDLgV+mHn9RwK70/PfD+wADmtoXwH8DDgZOCI9vmxoXcAfAY8Db2lRc8vt1qSmJcDfp/ui+qf5HHBMmvY7qc7TgMOB/w3cndqmAs8C56e6FqXH04b+Dt5ot54XMNZvKVh70x/ThhSiIxr+MP57w3NnAi8daE/TFgF3pfv/B/jjhrazMmFfBvxppqamYaf6R7IfmNzQfinwnXR/CXBHQ9tbgRczr//DVP8sxqfgPAcsbGhfAXyu4fGfALcNqeszwCPA8S1qzm63JjUtAV5OtbyQXu/FDe1XAV9peDwJeCWt83zgR0OWdz9wwdDfwRvt5m9Ch+e8iLijRdtTDfdPpNq7b5F0YNohDc85bsjzN2TWOYdqj3mwjgOeiYg9Q9bT+Fb96Yb7LwATJI2PiH1NlrcYuD617ZO0NE27KbO8oV9g/hnVP8WNLWqu227NXB8RH4bqm33gVkm7IuJbVNvggQNPjIi9knYCs1Pb0O2+IbW9oTnsnWscNvgU1R5qeovgbKEK8QEnZJb7FA2fQTPrHGozMFXS5IbAn0D1keCgSDqe6i3x/IZvu4+k+ucwPSJ2DHNRZwG3SXo6Im5s0l633bIiYlDSD4DfpfoosJnqH8iB1zERmEa1DV7TlpwA3HZgcQe7/tcLf0HXRRGxBVgOfFXSUZIOkXSSpP+QnnI98ClJx0uaAlySWdyVwGck/Ub6pv8tkg78kW4F3tyihqeA+4BLJU2Q9A7gI8C1bbyk84FHgV+j+uw/j+qz+Uaqt9nDtQ44G/hrSe9rUnPddstK/5TOTusB+B7wR5LmSToc+J/AqogYpPqS8WRJfyBpvKTfp/ooc2uat+W2fb1z2LvvD4HDqD6jPgvcAMxKbd+m+iz+ENXbzKWtFhIR/wR8meoPdw/VF1hTU/OlwOckPSfpM01mX0T1+XQz1dvtL0TE7W28lsXANyPi6cYbcDlDvpWvExEPAe8Fvi1pQZOn5LZbM79/oJ8dWA2sBL6Y1nUn8N+AG6neTZ0EfCi17Ux1fBrYCVwMvLfhXcrXgQ9IelbSXx3MaxzrlL6UMLM3OO/ZzQrhsJsVwmE3K4TDblaIUe1nnz59evT19Y3mKs2KMjg4yI4dO9SsraOwSzqbqqtiHHBlRFyWe35fXx8DAwOdrNLMMvr7W49pavttfBqR9NfAAqqDEhZJemu7yzOzkdXJZ/b5wOMR8UREvAz8A3Bud8oys27rJOyzee1AhY00GUwg6UJJA5IGtm/f3sHqzKwTnYS92ZcAv3I4XkRcERH9EdE/Y8aMDlZnZp3oJOwbee0IruOpjsU2szGok7CvBuZKepOkw6gGGtzSnbLMrNva7nqLiH2SPkk1imsccHVErKuZzcx6pKN+9oj4PtX4YDMb43y4rFkhHHazQjjsZoVw2M0K4bCbFcJhNyuEw25WCIfdrBAOu1khHHazQjjsZoVw2M0K4bCbFcJhNyuEw25WCIfdrBAOu1khHHazQjjsZoVw2M0K4bCbFcJhNyuEw25WCIfdrBAOu1khHHazQjjsZoVw2M0K4bCbFcJhNytER5dsljQI7AH2A/sior8bRZlZ93UU9uTdEbGjC8sxsxHkt/Fmheg07AEsl/RjSRc2e4KkCyUNSBrYvn17h6szs3Z1GvbTI+I0YAHwCUnvGvqEiLgiIvojon/GjBkdrs7M2tVR2CNic/q5DbgJmN+Nosys+9oOu6SJkiYfuA+cBaztVmFm1l2dfBs/E7hJ0oHlfC8ibutKVWbWdW2HPSKeAP5dF2sxsxHkrjezQjjsZoVw2M0K4bCbFcJhNytENwbCmPXE/v37s+2HHNJ6X5a6jNv20ksvZdsPP/zwbPtjjz3Wsm3u3Llt1VTHe3azQjjsZoVw2M0K4bCbFcJhNyuEw25WCIfdrBDuZy9cRHTUnuvLBti0aVPLtvvvvz8774IFC7LtEydOzLaPpLp+9DpLly5t2fbZz362o2W34j27WSEcdrNCOOxmhXDYzQrhsJsVwmE3K4TDblYI97NbVl0/ep177rmnZduqVauy827evDnb/qlPfaqtmrph27Zt2fZly5Zl2ydPntzNcobFe3azQjjsZoVw2M0K4bCbFcJhNyuEw25WCIfdrBDuZy9c3bnXx4/P/4msXr06275+/fqWbTNnzszOmzu3OsDChQuz7VOmTGnZ9otf/CI774knnpht37lzZ7Z99+7d2fbZs2dn20dC7Z5d0tWStkla2zBtqqTbJT2WfrbeqmY2Jgznbfx3gLOHTLsEuDMi5gJ3psdmNobVhj0i7gaeGTL5XOCadP8a4Lwu12VmXdbuF3QzI2ILQPp5bKsnSrpQ0oCkge3bt7e5OjPr1Ih/Gx8RV0REf0T0z5gxY6RXZ2YttBv2rZJmAaSf+SFAZtZz7Yb9FmBxur8YuLk75ZjZSKntZ5d0HXAGMF3SRuALwGXA9ZI+AjwJfHAki7T2vfrqq9n2un70559/Ptt+ww03ZNtz51ev6+ves2dPtr2Tc97Xzbtu3bps+/HHH59tz/XxQ/3xDSOhNuwRsahF03u6XIuZjSAfLmtWCIfdrBAOu1khHHazQjjsZoXwENdhynXVSMrOW9f9VTd/XXuuG2fcuHHZeetcfvnl2fa6YaoTJkxo2bZhw4bsvHVdc3Xr3rdvX8u2um1adznouks279q1K9v+0ksvtWyr6+5s91LV3rObFcJhNyuEw25WCIfdrBAOu1khHHazQjjsZoUopp+9bkhjp33dOZ1e9rhuOGQnfenXXXddtv3pp5/Otp966qnZ9lxf93PPPZedd+rUqdn2adOmZdt37NjRsm3v3r3ZeXN1D0fd39sLL7zQsq3uFNrz5s1rqybv2c0K4bCbFcJhNyuEw25WCIfdrBAOu1khHHazQhTTz95JPznkx6TXjVev6wevq62TfvSrr7462/7oo49m2+fMmZNtr7t0ca6/+cUXX8zOW3dZ47pTTee265FHHpmdt24sfafHbeQsW7Ys2+5+djPLctjNCuGwmxXCYTcrhMNuVgiH3awQDrtZIV5X/ex1/dk5df2edf2muTHpnY5Xr7N58+Zs+9KlS1u21fVlz507N9teN+47d/5zyPfDH3roodl5635nuTHhdep+Z3Xnha+bv+7c7rnXtnLlyuy87ar9K5V0taRtktY2TFsiaZOkB9PtnBGpzsy6Zji7pO8AZzeZ/rWImJdu3+9uWWbWbbVhj4i7gWdGoRYzG0GdfNj8pKSH09v8Ka2eJOlCSQOSBrZv397B6sysE+2G/W+Ak4B5wBbgq62eGBFXRER/RPTPmDGjzdWZWafaCntEbI2I/RHxKvBtYH53yzKzbmsr7JJmNTxcCKxt9VwzGxtq+9klXQecAUyXtBH4AnCGpHlAAIPAx4e7wk6uJT6S/dmdjD+u+y5icHAw2/7Tn/40275ly5Zs+2GHHday7aijjsrOW3fu9t27d2fbX3nllWx7rh++7vddt93qzu1+zDHHtGzLbTOoP1d/3XEZRxxxRNvLnzRpUnbetWtb71tzx1XUhj0iFjWZfFXdfGY2tvhwWbNCOOxmhXDYzQrhsJsVwmE3K8SoD3Ht5LTIW7dubdm2YcOG7LzPP/98R+25Lo2f//zn2XnrhmKOH5//NUyePDnbnhv6u2vXruy8dUNg62qre225Lqi6YaQvv/xytn3WrFnZ9ly3YV3dU6a0PAIcqB/6+8wz+eEkue61ustk55ad69Lznt2sEA67WSEcdrNCOOxmhXDYzQrhsJsVwmE3K8SYOpX0HXfckW3PnVK5rj+4bhhq3ZDG3PEBnfaT1/XZ1vW75oZb1p3qua4/ue703XW157Zr3emW64Z65oawQv3vvBN1261uOHbu+Ia64wvq/t5a1tTWXGb2uuOwmxXCYTcrhMNuVgiH3awQDrtZIRx2s0KMaj/77t27Wb58ecv2q67Kn7T2lFNOadlWN7a5kzHhkD/1cKenHa6rra7fNdenu2fPnuy8dbXVjXevOwV3btvUHT+QO38BwCOPPJJtz223ut9ZnbpjAOrOjzBhwoS2l33ssce2bMtdBtt7drNCOOxmhXDYzQrhsJsVwmE3K4TDblYIh92sEMO5ZPMc4LvAvwFeBa6IiK9Lmgr8I9BHddnm34uIZ3PLmjhxIvPnz2/Z/sMf/jBby5o1a1q23Xvvvdl56+T6JyHfFz516tTsvHXtRx99dLa9rp8911e+c+fO7Lx1l4uuO7963SWdc/3wDz30UHbed7zjHdn2vr6+bPvtt9/esq1unH+nlwevG3N+3HHHtWyru8x27tiJTs8bvw/4dET8W+DfA5+Q9FbgEuDOiJgL3Jkem9kYVRv2iNgSEQ+k+3uA9cBs4FzgmvS0a4DzRqpIM+vcQb1XkdQHnAqsAmZGxBao/iEArY/hM7OeG3bYJU0CbgQuioj8B7XXznehpAFJAzt27GinRjPrgmGFXdKhVEG/NiKWpslbJc1K7bOAbc3mjYgrIqI/IvqnT5/ejZrNrA21YVf1depVwPqI+MuGpluAxen+YuDm7pdnZt0ynCGupwPnA2skPZim/TlwGXC9pI8ATwIfrFvQuHHjsqf//fznPz+McpqrO6XxqlWrsu11XVD33Xdfy7bBwcHsvA8//HC2vW44ZN0w1Fz3Vl0XUl234Nvf/vZs+5lnnpltP+ecc1q25YZ5dsP73ve+lm1PPvlkdt5p06Zl2+u6x+qGLee65uouZX3yySe3bMtt09qwR8S9QKu/pvfUzW9mY4OPoDMrhMNuVgiH3awQDrtZIRx2s0I47GaFUF0fbjf19/fHwMDAqK3PrDT9/f0MDAw07Sr3nt2sEA67WSEcdrNCOOxmhXDYzQrhsJsVwmE3K4TDblYIh92sEA67WSEcdrNCOOxmhXDYzQrhsJsVwmE3K4TDblYIh92sEA67WSEcdrNCOOxmhXDYzQrhsJsVwmE3K0Rt2CXNkXSXpPWS1kn60zR9iaRNkh5Mt9YX4jaznqu9PjuwD/h0RDwgaTLwY0m3p7avRcRfjFx5ZtYttWGPiC3AlnR/j6T1wOyRLszMuuugPrNL6gNOBValSZ+U9LCkqyVNaTHPhZIGJA1s3769o2LNrH3DDrukScCNwEURsRv4G+AkYB7Vnv+rzeaLiCsioj8i+mfMmNGFks2sHcMKu6RDqYJ+bUQsBYiIrRGxPyJeBb4NzB+5Ms2sU8P5Nl7AVcD6iPjLhumzGp62EFjb/fLMrFuG82386cD5wBpJD6Zpfw4skjQPCGAQ+PiIVGhmXTGcb+PvBZpd7/n73S/HzEaKj6AzK4TDblYIh92sEA67WSEcdrNCOOxmhXDYzQrhsJsVwmE3K4TDblYIh92sEA67WSEcdrNCOOxmhVBEjN7KpO3AhoZJ04Edo1bAwRmrtY3VusC1taubtZ0YEU3P/zaqYf+VlUsDEdHfswIyxmptY7UucG3tGq3a/DberBAOu1kheh32K3q8/pyxWttYrQtcW7tGpbaefmY3s9HT6z27mY0Sh92sED0Ju6SzJf1U0uOSLulFDa1IGpS0Jl2GeqDHtVwtaZuktQ3Tpkq6XdJj6WfTa+z1qLYxcRnvzGXGe7rten3581H/zC5pHPAo8B+BjcBqYFFEPDKqhbQgaRDoj4ieH4Ah6V3AXuC7EfG2NO0rwDMRcVn6RzklIj47RmpbAuzt9WW809WKZjVeZhw4D7iAHm67TF2/xyhst17s2ecDj0fEExHxMvAPwLk9qGPMi4i7gWeGTD4XuCbdv4bqj2XUtahtTIiILRHxQLq/BzhwmfGebrtMXaOiF2GfDTzV8HgjY+t67wEsl/RjSRf2upgmZkbEFqj+eIBje1zPULWX8R5NQy4zPma2XTuXP+9UL8Le7FJSY6n/7/SIOA1YAHwivV214RnWZbxHS5PLjI8J7V7+vFO9CPtGYE7D4+OBzT2oo6mI2Jx+bgNuYuxdinrrgSvopp/belzPL42ly3g3u8w4Y2Db9fLy570I+2pgrqQ3SToM+BBwSw/q+BWSJqYvTpA0ETiLsXcp6luAxen+YuDmHtbyGmPlMt6tLjNOj7ddzy9/HhGjfgPOofpG/mfAf+1FDS3qejPwULqt63VtwHVUb+teoXpH9BFgGnAn8Fj6OXUM1fZ3wBrgYapgzepRbe+k+mj4MPBgup3T622XqWtUtpsPlzUrhI+gMyuEw25WCIfdrBAOu1khHHazQjjso0zSQkkh6ZRhPPcCScd1sK4zJN3a7brS8wclTW8yfe9B1tg4ynCNpLbGSUg6RtKftDNvKRz20bcIuJfqYKI6FwBth/0gHUxd3fbuiJgHfAD4qzaXcQzgsGc47KMoHRN9OtUBKB8a0nZx2rM9JOkySR8A+oFr017viMa9qaR+SSvS/fmS7pP0k/Tz17pRV3pnsELSDZL+VdK16SiwxnmPkHSbpI81We6fSVqdBnh8cRilHAU82zD/f5G0Nt0uqpl+GXBS2lb/62BefzF6cYRTqTfgw8BV6f59wGnp/oL0+Mj0eGr6uYJqbP2B+QeB6el+P7Ai3T8KGJ/unwncmO6fAdza8PwrD7KuM4BdVOMXDgHuB97ZUEsfcAfwhw3L2pt+nkV1IkWleW8F3tVk3YNUR4+tBV4A3pum/0aaPhGYRHVE46mZ6X3A2l7/jsfyzXv20bWIavw+6eeidP9M4G8j4gWAiDjYceJHA/+k6qwxXwN+fegTImIgIj56kHUB/CgiNkY1SONBqlAdcHOq+7tNlnlWuv0EeAA4BZjbYv3vjuoEGG8HvpHeabwTuCkino+IvcBS4Lcz063G+F4XUApJ04DfAd4mKYBxQEi6mGrvN5zjlvfx/z96TWiY/iXgrohYmMZJr+hSXQAvNTx9P6/9m1kJLJD0vUi748ZFA5dGxLeGW0tE/EzSVuCtNB8KTWa61fCeffR8gOoUTidGRF9EzAF+TrWnWg78Z0lHQnWutDTPHmBywzIGqd7GAry/YfrRwKZ0/4Iu1lXn88BO4JtN2pZRvaZJAJJmS8qeLCK1v4nqeoB3A+dJOjKNQFwI3JOZPnRb2RAO++hZRDU+vtGNwB9ExG1Uo50GJD0IfCa1fwe4/MAXdMAXga9LuodqL3vAV4BLJa2k2jP/ivSF3pUHU9cwX9dFwARV58b7pYhYDnwPuF/SGuAGWofxrvS67wIuiWp89wNUr/9HVGdzuTIifpKZvhNYmb608xd0TXjUm1khvGc3K4TDblYIh92sEA67WSEcdrNCOOxmhXDYzQrx/wApzdGAGJVoQwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAEWCAYAAACg3+FOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAXG0lEQVR4nO3de5Bc5X3m8e+DLmh0l5CE5JGQbBA2AoKAiZING0JirdewiYGKnUJbi/EuWLiA2mQrjuN1kjW7IYZKxXZcZa+NbBME8WWt2F6TrCFgBbBxFuMxESAsJxYgA0KXAQGSrRuSfvvHOeNtDdPvmenLdEvv86mamu7+9en+Tdc8fU73e95zFBGY2fHvhE43YGZjw2E3y4TDbpYJh90sEw67WSYcdrNMOOzHAElLJIWk8eX1uyVd1cDjnCLpp5LGtb5L63YOe4tI2iJpXxmmHZL+StLUdjxXRFwcEWtH2NPKmuWejYipEXG4Vb3UvIEM/oSkn9Vc/9VWPZc1x2Fvrd+KiKnAecAvAn889A4qHDeve80byNTybwc4p+a27wxdphu2LAa3knJy3PzTdZOI2ArcDZwFIOkBSX8m6bvAXuBNkmZI+rykbZK2SrppMASSxkn6C0kvSnoa+He1j18+3jU1198raZOkPZJ+KOk8SXcCpwB/W65hPzDMx4E3SLpL0i5JmyW9t+Yxb5T0FUl3lI/7pKS+Rl4PSX8t6VOS7pH0M+BXJc0sbx8ot0D+qySV979J0u01y58mKWquX10us0fS05KuqKldI+lHkl4uP+4sKm8fX/7t10naDPyokb/lmBYR/mnBD7AFWFleXgQ8Cfxpef0B4FngTGA8MAH438CtwBRgHvAIcG15//dR/DMuAmYD9wMBjK95vGvKy+8CtlJsSQg4DVg8tKfy+pIhj/Mg8D+BScByYAB4a1m7EdgPXAKMA24GHh7B6xDAaUNu+2vgZeBfUaxgTgS+CHwNmAa8CdgMXFXe/ybg9prlTyv+VQNgOvAqsLS8vgBYVl5+J/DPwJvL1/lG4DtlbXzZ2z3ALKCn0/8zY/4/2ukGjpefMlg/BV4BflKGqKesPQD8j5r7ngwcqP2HA1YB95eX/wF4X03tbYmw/z3wu4mehg17+UZyGJhWU795MGRlUL5VU1sG7BvB61Av7LfVXJ8AHAJOr7nt+sHnG0HYXwEuByYNeZ77Bt8wyuvjy9e5tybsF3b6f6VTP96Mb63LImJmRCyOiOsiYl9N7bmay4sp/uG3SXpF0isUa/l5Zf0NQ+7/k8RzLgKeaqDXNwC7ImLPkOfprbm+vebyXmBSE591a/+eeRRbC7V/19DnHlZE7KZ4Y7we2C7p7ySdXpYXA5+qeU1fBI4AC+v0kRWHfezUTi98jmKNM6d8c5gZEdMj4syyvo0ixINOSTzuc8CpI3jOoV4AZkuaNuR5tiaWaUZtLzsptioW13nunwGTa2rzj3qgiLsjYiXFJvxmijdKKF6Lq2te05kR0RMR36vTR1Yc9g6IiG3AvcBHJU2XdIKkUyX9WnmXrwD/WdJCSbOADyYe7nPA+yWdX37Tf5qkwRDtoPg8PFwPzwH/CNwsaZKkXwCuBr7Qgj8xKSJeA/4G+IikqZLeCPwXis19gA3Ar0laJGkmNX+/pAWSfkvSZOAgxRvD4FDiZ4A/knRGed+Zkt7Z7r/nWOGwd867gYnADym+vPobijUVwGcpPos/BjxK8UXWsCJiHfBnFF947aH44m92Wb4Z+ONys/b9wyy+iuJz/AvA14EPR8R9Tf1VI3cdRVifofiicC1wR1m7p+znCYovLu+qWW4c8AcUWz8vAb8C3AA/fy0+BqyTtBt4HPi37f5DjhUqv8gws+Oc1+xmmXDYzTLhsJtlwmE3y8SYTgaYM2dOLFmyZCyf8phw6NChZH1gYCBZHzeu/rySE05o7v089djNqvpyePz49L/ntGnTkvVyV/usbNmyhRdffHHYP7ypsEt6O/AJiuGQz0XELan7L1myhP7+/mae8rhUFeZbb701WZ85c2bdWk9PT0M9DZoxY0ayXhWow4frz6Y9ePBgctl58+Yl6xdddFGyPnHixGT9eNTXV3+uUsNv++UMrU8BF1PsN71K0rJGH8/M2quZbbwVwOaIeDoiDgJfBi5tTVtm1mrNhL2XoycVPM8wExkkrZbUL6m/anPVzNqnmbAP92Htdd+4RMSaiOiLiL65c+c28XRm1oxmwv48R8/MWkixj7WZdaFmwv59YKmkN0qaCFzB0RMWzKyLNDz0FhGHJN1AMTtrHMWRSJ5sWWcZWbduXbJ+0003JeuzZs2qW1uwYEHdGsAzzzyTrPf2po8ncfrppyfrmzZtqlubNGlSctmVK1cm6zt27EjWr7zyymQ9N02Ns0fEN4FvtqgXM2sj7y5rlgmH3SwTDrtZJhx2s0w47GaZcNjNMpHdye26UdWcgapjADQzZ33+/PnJemqKKsBLL72UrL/66qt1a9OnT08uu3Vr+hD2b3nLW5J1O5rX7GaZcNjNMuGwm2XCYTfLhMNulgmH3SwTHnrrAlXDV1VH+HnqqfqnZ589e3bdGsCePXuS9arDNb/yyivJeupw0bt3704uWzWkePbZZyfrdjSv2c0y4bCbZcJhN8uEw26WCYfdLBMOu1kmHHazTHicvQssXrw4WX/ssceS9dRplatOuTxlypRkvepMqFVTYFNTaF9++eXkskeOHEnWPcV1dLxmN8uEw26WCYfdLBMOu1kmHHazTDjsZplw2M0y4XH2LiApWa+at50aK0/NJ4f0XHhofiy86pTOKaeeemqyPn68/31Ho6lXS9IWYA9wGDgUEX2taMrMWq8Vb42/HhEvtuBxzKyN/JndLBPNhj2AeyX9QNLq4e4gabWkfkn9Vac5MrP2aTbsF0TEecDFwPWSLhx6h4hYExF9EdFXdeBEM2ufpsIeES+Uv3cCXwdWtKIpM2u9hsMuaYqkaYOXgbcBG1vVmJm1VjPfxp8MfL0cIx4PfDEi7mlJV5mpOj76okWLkvVly5bVrVWN4a9bty5Z37VrV7L+5JNPJusXXvi6T3Y/d/755yeX7e3tTdYPHjyYrE+ePDlZz03DYY+Ip4FzWtiLmbWRh97MMuGwm2XCYTfLhMNulgmH3SwTniPYBc4444xkff369Q0vf+KJJyaXPfPMM5P1FSvS+0mtXj3sXtI/d8opp9StLVy4MLnsrFmzkvWenp5k3Y7mNbtZJhx2s0w47GaZcNjNMuGwm2XCYTfLhMNulgmPs3eBvXv3JutVp1Xevn173VrVWHWVQ4cOJesHDhxI1lOHmp40aVJy2apDRe/fvz9Zr9rHIDdes5tlwmE3y4TDbpYJh90sEw67WSYcdrNMOOxmmfA4exeoGkevGodPHYr6hRdeSC5bNY6+fPnyZL3qUNX79u2rW6s6FPThw4eT9QkTJiTrdjSv2c0y4bCbZcJhN8uEw26WCYfdLBMOu1kmHHazTHicvQtUHf+86pTOU6dObfi5q5Y999xzG35sSO8jUPV3V81H9zj76FSu2SXdJmmnpI01t82WdJ+kH5e/mztCgpm13Ug2428H3j7ktg8C6yNiKbC+vG5mXawy7BHxbWDXkJsvBdaWl9cCl7W4LzNrsUa/oDs5IrYBlL/n1bujpNWS+iX1DwwMNPh0Ztastn8bHxFrIqIvIvrmzp3b7qczszoaDfsOSQsAyt87W9eSmbVDo2G/C7iqvHwV8I3WtGNm7VI5zi7pS8BFwBxJzwMfBm4BviLpauBZ4F3tbPJ4N27cuGS9ajw5Nae8ar55M2P0UD1W/tprr9WtVc3jr3pdqup2tMqwR8SqOqW3trgXM2sj7y5rlgmH3SwTDrtZJhx2s0w47GaZ8BTXLjBnzpxkvWr4LHVa5KrDNVedNrnKtGnTkvWIaPi5e3t7k/Wqqb92NL9aZplw2M0y4bCbZcJhN8uEw26WCYfdLBMOu1kmPM7eBRYsWJCsV42Vp8ayq073nJqCOhJVp3xOTWOdPn16ctnU/gM2el6zm2XCYTfLhMNulgmH3SwTDrtZJhx2s0w47GaZ8Dh7F5g8eXJT9dThoKvGqnftGnoav9GpOhz0gQMH6taqTsl80kknNdSTDc9rdrNMOOxmmXDYzTLhsJtlwmE3y4TDbpYJh90sEx5n7wJVpx6uGstOjaVXzTefO3dusl5l6dKlyfq+ffvq1qrm0u/fv7+hnmx4lWt2SbdJ2ilpY81tN0raKmlD+XNJe9s0s2aNZDP+duDtw9z+8YhYXv58s7VtmVmrVYY9Ir4NNLdPpZl1XDNf0N0g6fFyM39WvTtJWi2pX1L/wMBAE09nZs1oNOyfBk4FlgPbgI/Wu2NErImIvojoa/bLIDNrXENhj4gdEXE4Io4AnwVWtLYtM2u1hsIuqfbYx5cDG+vd18y6Q+U4u6QvARcBcyQ9D3wYuEjSciCALcC1bewxe1Vj5ak56VXLzppV9+uWEVm2bFmy/txzz9Wt7d69O7lsT09PQz3Z8CrDHhGrhrn5823oxczayLvLmmXCYTfLhMNulgmH3SwTDrtZJjzF9Rjw0ksvJeupaaZ33313ctlrr21u1PS8885L1h955JG6td7e3uSyPmVza3nNbpYJh90sEw67WSYcdrNMOOxmmXDYzTLhsJtlwuPsx4AHH3wwWd+8eXPdWtU4+5133tlQT4POOuusZD01/faTn/xkctlzzjknWT///POTdTua1+xmmXDYzTLhsJtlwmE3y4TDbpYJh90sEw67WSY8zt4FIiJZP3z4cLKeGmevOqXypEmTkvUq48en/4VeffXVurXUXHeoPgy2jY7X7GaZcNjNMuGwm2XCYTfLhMNulgmH3SwTDrtZJkZyyuZFwB3AfOAIsCYiPiFpNvC/gCUUp23+nYh4uX2tHr8kJesHDx5M1vft21e3duKJJzbU00hV9fbaa6/VraXG4KuWtdEbyZr9EPD7EXEG8MvA9ZKWAR8E1kfEUmB9ed3MulRl2CNiW0Q8Wl7eA2wCeoFLgbXl3dYCl7WrSTNr3qg+s0taApwLfA84OSK2QfGGAMxrdXNm1jojDrukqcBXgd+LiN2jWG61pH5J/QMDA430aGYtMKKwS5pAEfQvRMTXypt3SFpQ1hcAO4dbNiLWRERfRPTNnTu3FT2bWQMqw67iq+LPA5si4mM1pbuAq8rLVwHfaH17ZtYqI5niegFwJfCEpA3lbR8CbgG+Iulq4FngXe1p0SZOnJis795d/1PVlClTWt3OUSZMmJCsp6bAVg2tzZ8/v6GebHiVYY+Ih4B6A8FvbW07ZtYu3oPOLBMOu1kmHHazTDjsZplw2M0y4bCbZcKHkj4G9PT0JOv79++vW2v2UNFVqvYBSB0m+8iRI009to2O1+xmmXDYzTLhsJtlwmE3y4TDbpYJh90sEw67WSY8zn4M2L59e7KeOqVz1Vh2s6ZOnZqsn3BC/fVJ1amoq/YvsNHxmt0sEw67WSYcdrNMOOxmmXDYzTLhsJtlwmE3y4TH2Y8BJ598crK+c+ewJ+MBYNy4ca1u5yizZs1K1lPj7AcOHEguO2+eTx/YSl6zm2XCYTfLhMNulgmH3SwTDrtZJhx2s0w47GaZqBxnl7QIuAOYDxwB1kTEJyTdCLwXGCjv+qGI+Ga7Gs3ZxRdfnKz39/fXrbV7nH3atGnJ+owZM+rWUse7B1i8eHFDPdnwRrJTzSHg9yPiUUnTgB9Iuq+sfTwi/qJ97ZlZq1SGPSK2AdvKy3skbQJ6292YmbXWqD6zS1oCnAt8r7zpBkmPS7pN0rD7TUpaLalfUv/AwMBwdzGzMTDisEuaCnwV+L2I2A18GjgVWE6x5v/ocMtFxJqI6IuIvrlz57agZTNrxIjCLmkCRdC/EBFfA4iIHRFxOCKOAJ8FVrSvTTNrVmXYJQn4PLApIj5Wc/uCmrtdDmxsfXtm1ioj+Tb+AuBK4AlJG8rbPgSskrQcCGALcG1bOrTK0y6nhrDaPfRWZd++fXVre/fuTS67cOHCVreTtZF8G/8QoGFKHlM3O4Z4DzqzTDjsZplw2M0y4bCbZcJhN8uEw26WCR9K+hjw7ne/O1l/6KGH6taqpse22zve8Y6Glz377LNb2Il5zW6WCYfdLBMOu1kmHHazTDjsZplw2M0y4bCbZUIRMXZPJg0AP6m5aQ7w4pg1MDrd2lu39gXurVGt7G1xRAx7/LcxDfvrnlzqj4i+jjWQ0K29dWtf4N4aNVa9eTPeLBMOu1kmOh32NR1+/pRu7a1b+wL31qgx6a2jn9nNbOx0es1uZmPEYTfLREfCLuntkv5Z0mZJH+xED/VI2iLpCUkbJNU/F/LY9HKbpJ2SNtbcNlvSfZJ+XP4e9hx7HertRklby9dug6RLOtTbIkn3S9ok6UlJv1ve3tHXLtHXmLxuY/6ZXdI44F+AfwM8D3wfWBURPxzTRuqQtAXoi4iO74Ah6ULgp8AdEXFWedufA7si4pbyjXJWRPxhl/R2I/DTTp/Guzxb0YLa04wDlwHvoYOvXaKv32EMXrdOrNlXAJsj4umIOAh8Gbi0A310vYj4NrBryM2XAmvLy2sp/lnGXJ3eukJEbIuIR8vLe4DB04x39LVL9DUmOhH2XuC5muvP013new/gXkk/kLS6080M4+SI2AbFPw8wr8P9DFV5Gu+xNOQ0413z2jVy+vNmdSLsw51KqpvG/y6IiPOAi4Hry81VG5kRncZ7rAxzmvGu0Ojpz5vVibA/Dyyqub4QeKEDfQwrIl4of+8Evk73nYp6x+AZdMvfOzvcz89102m8hzvNOF3w2nXy9OedCPv3gaWS3ihpInAFcFcH+ngdSVPKL06QNAV4G913Kuq7gKvKy1cB3+hgL0fpltN41zvNOB1+7Tp++vOIGPMf4BKKb+SfAv6oEz3U6etNwGPlz5Od7g34EsVm3WsUW0RXAycB64Efl79nd1FvdwJPAI9TBGtBh3r71xQfDR8HNpQ/l3T6tUv0NSavm3eXNcuE96Azy4TDbpYJh90sEw67WSYcdrNMOOxdRtLlkkLSW0Zw3/dIekMTz3WRpL9L1E+qmYm1fcjMrImNPq91hsPefVYBD1HsbFTlPUDDYa8SES9FxPKIWA58Bvj44PUoJjEBxc4iksbsf0mSTzXeAIe9i5T7TF9AsYPKFUNqHyjn2T8m6RZJ7wT6gC+Ua9qeci7+nPL+fZIeKC+vkPSPkv6p/P3mFvR6mqSNkj4DPAoskPQfyh43SvpIeb/xkl6pWe4KSZ+rubyx/Jvur7n/xyQ9Uk4Muaa8faWkb0n6MvBPzfafI79DdpfLgHsi4l8k7ZJ0XhRzny8ua78UEXslzY6IXZJuAN4fEf0Axd6Yw/oRcGFEHJK0EvgI8Nu1d5DUB7wvIq4ZRb/LgP8YEe+TtBC4ieIN6FXgW5J+E7gnsfyHgYsiYoekmeVtq4GdEbFC0onAw5LuLWu/DCyLiGdH0aOVHPbusgr4y/Lyl8vrjwIrgb+KiL0AETHaeeQzgLWSllLsrjlh6B3KN4zRBB3gqYj4fnn5l4B/iPKgH5K+CFxIOuzfBe6QtA4YnKzyNuAMSYNbNjOApeXl/+ugN85h7xKSTgJ+AzhLUgDjgJD0AYppwSPZr/kQ//+j2aSa2/8UuD8iLi/nUT/QorZ/VnO53mbFkSG12r7eS/Em8ZvAY5J+obzvdRGxvvZByi2S2uezUfJn9u7xTopDPC2OiCURsQh4hmLyxL3Af5I0GYpjqZXL7AGm1TzGFuD88nLtZvoMYGt5+T1t6R4eBn69/AZ/PMV3Dg9GMW3zZUlLyy/xLq9Z5k0R8TDwJ8DLFAcx+XvgusEv4SS9WVJPm3rOisPePVZRzJ+v9VXg30fEPRSzofolbQDeX9ZvBz4z+AUd8N+BT0j6DnC45nH+HLhZ0ncpthhep/xC73ONNh8RzwP/jWKrYQPwcET8n7L8hxSb8+spZsgN+rikJyhmfH0rIjYCt1LMStug4mCWn8ZboC3hWW9mmfCa3SwTDrtZJhx2s0w47GaZcNjNMuGwm2XCYTfLxP8DLQomzuJ8iCAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#model evaluation\n",
    "test_loss, test_acc = model.evaluate(test_images, test_labels)\n",
    "print(\"Accuracy: \", test_acc)\n",
    "\n",
    "prediction = model.predict(test_images)\n",
    "#displaying predictions\n",
    "for i in range(7):\n",
    "    plt.grid(False)\n",
    "    plt.imshow(test_images[i], cmap=plt.cm.binary)\n",
    "    plt.xlabel(\"Actual: \" + class_names[test_labels[i]])\n",
    "    plt.title(\"Prediction \" + class_names[np.argmax(prediction[i])])\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}
À propos de cet Algorithme
#Importing Packages
from tensorflow import keras 
import numpy as np           
import matplotlib.pyplot as plt
#Importing Keras Dataset
data = keras.datasets.fashion_mnist

#train and test data segregation
(train_images, train_labels), (test_images, test_labels) = data.load_data()

class_names = ["T-shirt/top", "Trouser", "Pullover", "Dress", "Coat", "Sandal", "Shirt", "Sneaker", "Bag", "Ankle Boot"]

train_images = train_images/255.0
test_images = test_images/255.0
Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz
32768/29515 [=================================] - 0s 3us/step
Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz
26427392/26421880 [==============================] - 25s 1us/step
Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz
8192/5148 [===============================================] - 0s 0us/step
Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz
4423680/4422102 [==============================] - 4s 1us/step
#Create a object of model class
model = keras.Sequential([
    keras.layers.Flatten(input_shape=(28,28)),
    keras.layers.Dense(128, activation="relu"),
    keras.layers.Dense(10, activation="softmax")
])
#Compile the model with MSE loss and Adam optimizer
model.compile(optimizer="adam", loss="sparse_categorical_crossentropy", metrics=["accuracy"])
#fitting the model
model.fit(train_images, train_labels, epochs=5)
WARNING:tensorflow:From C:\Users\vinay\Anaconda3\lib\site-packages\tensorflow\python\ops\init_ops.py:1251: calling VarianceScaling.__init__ (from tensorflow.python.ops.init_ops) with dtype is deprecated and will be removed in a future version.
Instructions for updating:
Call initializer instance with the dtype argument instead of passing it to the constructor
Epoch 1/5
60000/60000 [==============================] - 8s 132us/sample - loss: 0.4980 - acc: 0.8253
Epoch 2/5
60000/60000 [==============================] - 8s 127us/sample - loss: 0.3717 - acc: 0.8662
Epoch 3/5
60000/60000 [==============================] - 6s 93us/sample - loss: 0.3341 - acc: 0.8787
Epoch 4/5
60000/60000 [==============================] - 8s 130us/sample - loss: 0.3126 - acc: 0.8852
Epoch 5/5
60000/60000 [==============================] - 6s 92us/sample - loss: 0.2928 - acc: 0.8929
&lt;tensorflow.python.keras.callbacks.History at 0x25f99983fd0&gt;
#model evaluation
test_loss, test_acc = model.evaluate(test_images, test_labels)
print("Accuracy: ", test_acc)

prediction = model.predict(test_images)
#displaying predictions
for i in range(7):
    plt.grid(False)
    plt.imshow(test_images[i], cmap=plt.cm.binary)
    plt.xlabel("Actual: " + class_names[test_labels[i]])
    plt.title("Prediction " + class_names[np.argmax(prediction[i])])
    plt.show()
10000/10000 [==============================] - 1s 79us/sample - loss: 0.3884 - acc: 0.8587
Accuracy:  0.8587