{"cells":[{"metadata":{},"cell_type":"markdown","source":"# This is my first kernel for the MNIST Digit Recognizer Competition\n## All help and feedback is welcome.\n\nCredits to this kernel [here](https://www.kaggle.com/dejavu23/mnist-sklearn-and-keras)"},{"metadata":{},"cell_type":"markdown","source":"***"},{"metadata":{},"cell_type":"markdown","source":"## Loading relevant models in the beggining of the code\n#### I am also asserting that we re using python 3.5 and scikit-learn version 0.20 "},{"metadata":{},"cell_type":"markdown","source":"***"},{"metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","trusted":true},"cell_type":"code","source":"# Python ≥3.5 is required\nimport sys\nassert sys.version_info >= (3, 5)\n\n# Scikit-Learn ≥0.20 is required\nimport sklearn\nassert sklearn.__version__ >= \"0.20\"\n\n# Common imports\nimport numpy as np\nimport os\n\n# To plot pretty figures\nimport matplotlib as mpl\nimport matplotlib.pyplot as plt\nmpl.rc('axes', labelsize=14)\nmpl.rc('xtick', labelsize=12)\nmpl.rc('ytick', labelsize=12)\n\nfrom sklearn.pipeline import make_pipeline\nfrom sklearn.model_selection import RandomizedSearchCV, GridSearchCV\n\nimport pandas as pd\nimport seaborn as sns #for better and easier plots\n\n%matplotlib inline\n\n\nimport warnings\nwarnings.filterwarnings(action=\"ignore\")","execution_count":1,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"#### loading our training and test set"},{"metadata":{"_cell_guid":"79c7e3d0-c299-4dcb-8224-4455121ee9b0","_uuid":"d629ff2d2480ee46fbb7e2d37f6b5fab8052498a","trusted":true},"cell_type":"code","source":"train = pd.read_csv(\"../input/train.csv\")\ntest = pd.read_csv(\"../input/test.csv\")","execution_count":2,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"train.head(3) #looking at the first 3 entries, just to have an overall picture of what we have at hand","execution_count":3,"outputs":[{"output_type":"execute_result","execution_count":3,"data":{"text/plain":"   label  pixel0  pixel1    ...     pixel781  pixel782  pixel783\n0      1       0       0    ...            0         0         0\n1      0       0       0    ...            0         0         0\n2      1       0       0    ...            0         0         0\n\n[3 rows x 785 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>label</th>\n      <th>pixel0</th>\n      <th>pixel1</th>\n      <th>pixel2</th>\n      <th>pixel3</th>\n      <th>pixel4</th>\n      <th>pixel5</th>\n      <th>pixel6</th>\n      <th>pixel7</th>\n      <th>pixel8</th>\n      <th>pixel9</th>\n      <th>pixel10</th>\n      <th>pixel11</th>\n      <th>pixel12</th>\n      <th>pixel13</th>\n      <th>pixel14</th>\n      <th>pixel15</th>\n      <th>pixel16</th>\n      <th>pixel17</th>\n      <th>pixel18</th>\n      <th>pixel19</th>\n      <th>pixel20</th>\n      <th>pixel21</th>\n      <th>pixel22</th>\n      <th>pixel23</th>\n      <th>pixel24</th>\n      <th>pixel25</th>\n      <th>pixel26</th>\n      <th>pixel27</th>\n      <th>pixel28</th>\n      <th>pixel29</th>\n      <th>pixel30</th>\n      <th>pixel31</th>\n      <th>pixel32</th>\n      <th>pixel33</th>\n      <th>pixel34</th>\n      <th>pixel35</th>\n      <th>pixel36</th>\n      <th>pixel37</th>\n      <th>pixel38</th>\n      <th>...</th>\n      <th>pixel744</th>\n      <th>pixel745</th>\n      <th>pixel746</th>\n      <th>pixel747</th>\n      <th>pixel748</th>\n      <th>pixel749</th>\n      <th>pixel750</th>\n      <th>pixel751</th>\n      <th>pixel752</th>\n      <th>pixel753</th>\n      <th>pixel754</th>\n      <th>pixel755</th>\n      <th>pixel756</th>\n      <th>pixel757</th>\n      <th>pixel758</th>\n      <th>pixel759</th>\n      <th>pixel760</th>\n      <th>pixel761</th>\n      <th>pixel762</th>\n      <th>pixel763</th>\n      <th>pixel764</th>\n      <th>pixel765</th>\n      <th>pixel766</th>\n      <th>pixel767</th>\n      <th>pixel768</th>\n      <th>pixel769</th>\n      <th>pixel770</th>\n      <th>pixel771</th>\n      <th>pixel772</th>\n      <th>pixel773</th>\n      <th>pixel774</th>\n      <th>pixel775</th>\n      <th>pixel776</th>\n      <th>pixel777</th>\n      <th>pixel778</th>\n      <th>pixel779</th>\n      <th>pixel780</th>\n      <th>pixel781</th>\n      <th>pixel782</th>\n      <th>pixel783</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>1</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>...</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n      <td>0</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"* our target is the label column, so I am gonna separate it and create our Y_value, which is our ground truth, the value that we want to predict"},{"metadata":{"trusted":true},"cell_type":"code","source":"train_labels = train['label'].copy()\ntrain.drop('label', axis=1, inplace=True)","execution_count":4,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(train.shape, test.shape) #checking the shape","execution_count":5,"outputs":[{"output_type":"stream","text":"(42000, 784) (28000, 784)\n","name":"stdout"}]},{"metadata":{},"cell_type":"markdown","source":"### the dataset is composed of 784 columns, each image is 28 pixels in height and 28 pixels in width, for a total of 784 pixels in total. "},{"metadata":{},"cell_type":"markdown","source":"* Normalization \n* dividing each pixel and label of our dataset by 255"},{"metadata":{"trusted":true},"cell_type":"code","source":"train /= 255\ntest /= 255","execution_count":6,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"#### Plotting the distribution of digits in our dataset"},{"metadata":{"trusted":true},"cell_type":"code","source":"fig, ax = plt.subplots(figsize=(8,5))\nsns.countplot(train_labels)","execution_count":7,"outputs":[{"output_type":"execute_result","execution_count":7,"data":{"text/plain":"<matplotlib.axes._subplots.AxesSubplot at 0x7f1cd24805c0>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x360 with 1 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAAgcAAAFHCAYAAAA1Ee/KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAF6lJREFUeJzt3X+QZWV95/H3R4aIMowBGQlrypmVoOCo4DKlW7EQXVRwtzTqmCwJ/l7FQGGZJWrMBnSCUlQ0iUZFDBaKxh8rGDT+KrIxGgyYTRyNYCYQNsRFUSbO6GSY4ceA+t0/zmnrPr0z032Z7ntuN+9X1S3ufZ7z3P7eqab70895znlSVUiSJM14wNAFSJKk6WI4kCRJDcOBJElqGA4kSVLDcCBJkhqGA0mS1DAcSJKkhuFAkiQ1DAeSJKmxYugChnL44YfX2rVrhy5DkqSJ+drXvratqlbPddz9NhysXbuWTZs2DV2GJEkTk+SW+RznaQVJktQwHEiSpIbhQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDXut3srLAffPv9xQ5fQeMQbvzl0CZKkBeDMgSRJahgOJElSw3AgSZIahgNJktQwHEiSpIbhQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqbFi6AIkSYtv48aNQ5fQmLZ61HLmQJIkNZw5kOZw9VNOGrqEnzrpy1cPXYKk+wFnDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqWE4kCRJDcOBJElqeBMkSZIWyHGf+POhS/ip615wyn0eazjQRD35XU8euoTGta++dugSJGnqeFpBkiQ1nDkYccLrPjR0CY2vve3FQ5cgaS9uuOCLQ5fwU8f+zn8augQtM84cSJKkhuFAkiQ1DAeSJKkxyJqDJEcD3wQ+UVUv7Nt+DbgQOBz4C+DlVfXDvu8w4FLgmcA24Ler6qMj77fXsdL9zbt/8zNDl9A4+w+ePXQJWqIuv+KJQ5fQ+JVf/ruhS5iYoWYOLgK+OvMiyTrgj4EXAUcAdwLvmXX8PX3f6cDF/Zj5jJUkSWOY+MxBktOAfwO+AvxC33w68Jmq+nJ/zHnADUkOAX4CbAAeW1W7gGuSfJouDLxhX2OraucEP5okScvCRGcOkqwCzgfOmdW1Drhu5kVV3Uw3U/Co/vGjqrpp5Pjr+jFzjZUkSWOa9GmFNwOXVtWts9pXAjtmte0ADun7bt9L31xjG0nOSLIpyaatW7feh/IlSVr+JhYOkhwPPB14+x66dwGrZrWtAnbO0TfX2EZVXVJV66tq/erVq8f7AJIk3U9Mcs3BU4G1wLeTQPcX/wFJHgNcBRw3c2CSRwIPBG6iW3OwIsnRVfV/+kOOAzb3zzfvY6wkSRrTJMPBJcD/HHn9WrqwcCbwMOBvkpwIfJ1uXcKVMwsKk1wJnJ/kFcDxwC8Bv9i/z0f2NVaSJI1nYqcVqurOqtoy86A7HXB3VW2tqs3Ar9P9ov8+3XqBs0aGnwU8qO/7GHBmP4Z5jJUkSWMYbOOlqto46/VHgY/u5dgfAs/dx3vtdawkSRqPt0+WJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUGOxqBUmaccELXzB0CY3f+fAnhi5BGpQzB5IkqWE4kCRJDcOBJElqGA4kSVLDcCBJkhqGA0mS1DAcSJKkhuFAkiQ1DAeSJKlhOJAkSQ3DgSRJahgOJElSw3AgSZIahgNJktQwHEiSpIbhQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqWE4kCRJDcOBJElqGA4kSVLDcCBJkhqGA0mS1DAcSJKkhuFAkiQ1DAeSJKlhOJAkSQ3DgSRJahgOJElSw3AgSZIahgNJktQwHEiSpIbhQJIkNSYaDpJ8OMltSW5PclOSV4z0nZzkxiR3JvlSkjUjfQ9M8v5+3JYk58x6372OlSRJ45n0zMGFwNqqWgU8B3hLkhOSHA5cCZwHHAZsAj4+Mm4jcDSwBnga8PokpwLMY6wkSRrDikl+saraPPqyfxwFnABsrqorAJJsBLYlOaaqbgReAry0qrYD25O8D3gpcBXw/DnGSpKkMUx8zUGS9yS5E7gRuA34PLAOuG7mmKq6A7gZWJfkUODI0f7++br++V7H7uFrn5FkU5JNW7duXdDPJUnScjHxcFBVZwGHACfSnQ7YDawEdsw6dEd/3MqR17P7mGPs7K99SVWtr6r1q1ev3p+PIUnSsjXI1QpV9eOqugb4eeBMYBewatZhq4CdfR+z+mf6mGOsJEka09CXMq6gW3OwGThupjHJwTPt/TqD20b7++cz6xf2OnZRK5ckaZmaWDhI8rAkpyVZmeSAJKcAvwr8JfBJ4LFJNiQ5CHgjcP3IgsIPAecmOTTJMcArgcv6vrnGSpKkMUxy5qDoTiHcCmwHfh/4jar6dFVtBTYAF/R9TwJOGxn7JrpFhrcAVwNvq6qrAOYxVpIkjWFilzL2v8RP2kf/F4Bj9tK3G3h5/xhrrCRJGs/Qaw4kSdKUMRxIkqSG4UCSJDUMB5IkqWE4kCRJjXmHgySPSJI9tCfJIxa2LEmSNJRxZg6+BexpQ4LD+j5JkrQMjBMOQncjo9lWAncvTDmSJGloc94EKck7+6cFXNhvtzzjAOCJwDcWoTZJkjSA+dwh8XH9fwMcC9wz0ncP8HW6WyFLkqRlYM5wUFVPA0jyAeA1VXX7olclSZIGM++9FarqZYtZiCRJmg7zDgf9dsivAU4GHsasxYxV9fiFLU2SJA1hnF0Z3wM8D7gC+Ap7vnJBkiQtceOEg+cCv9xvjyxJkpapce5zcCfwncUqRJIkTYdxwsFbgXP2dAtlSZK0fIxzWuEZwInAqUn+Ebh3tLOqnrOQhUmSpGGMEw62AZ9crEIkSdJ08D4HkiSpMc6aA0mSdD8wzk2Qvsk+7m3gTZAkSVoexllz8IlZrw8EjgeeDFy0YBVJkqRBjbPm4Hf31J7kdcCaBatIkiQNaiHWHFwJnL4A7yNJkqbAQoSDp9DdPVGSJC0D4yxI/PTsJuBI4AnAHk85SJKkpWecBYk/mPX6J8Bm4H9U1f9auJIkSdKQvAmSJElqjDNzAECSRwKPobvnwQ1V9S8LXpUkSRrMOGsOVgGXAhvoTin0zflT4L9V1c5FqE+SJE3YOFcr/BHweOBpwIP6x8l92zsWvjRJkjSEccLBc4BXVNXVVXVv//gr4AzguYtSnSRJmrhxwsGD+P+vWAD4IXDQwpQjSZKGNk44uBZ4c5IHzzQkOZjuHgdfWejCJEnSMMa5WuEc4Crgu0mu79seB9wFPHOhC5MkScMY5z4H30xyNN0+Csf0zX8CfKSq7lqM4iRJ0uSNcynjBcB3quq9s9p/PcnDq+q8Ba9OkiRN3DhrDl4E/P0e2r8OvHhhypEkSUMbJxw8DNi6h/ZtwBELU44kSRraOOHg28CJe2h/CnDrwpQjSZKGNs7VCn8MvD3JzwBf7NtOBi4Efm+hC5MkScMY52qFP0hyOPBO4Gf65nuAP6qqty5GcZIkafLG2pWxqn47yVvodmWEblfGXQtfliRJGsrYWzZX1R3AVxehFkmSNAXGWZC4X5I8MMmlSW5JsjPJN5I8a6T/5CQ3JrkzyZeSrJk19v1Jbk+yJck5s957r2MlSdJ4JhYO6GYpvgOcBDwEOBe4PMnafi3DlcB5wGHAJuDjI2M3AkcDa+i2jH59klMB5jFWkiSNYezTCvdVfzpi40jTZ5N8CzgBeCiwuaquAEiyEdiW5JiquhF4CfDSqtoObE/yPuCldHs9PH+OsZIkaQyTnDloJDkCeBSwGVgHXDfT1weJm4F1SQ4Fjhzt75+v65/vdewevuYZSTYl2bR1657u5yRJkgYJB0kOBD4CfLD/634lsGPWYTuAQ/o+ZvXP9DHH2EZVXVJV66tq/erVq/fvQ0iStExNPBwkeQDdbo73AGf3zbuAVbMOXQXs7PuY1T/TN9dYSZI0pomGgyQBLqXbi2FDVd3bd20Gjhs57mDgKLq1BNuB20b7++eb5xq7SB9DkqRlbdIzBxcDxwLPrqq7Rto/CTw2yYYkBwFvBK4fWVD4IeDcJIcmOQZ4JXDZPMdKkqQxTPI+B2uAVwHHA1uS7Oofp1fVVmADcAGwHXgScNrI8DfRLTK8BbgaeFtVXQUwj7GSJGkMk7yU8RYg++j/AnDMXvp2Ay/vH2ONlSRJ4xnsUkZJkjSdDAeSJKlhOJAkSQ3DgSRJahgOJElSw3AgSZIahgNJktQwHEiSpIbhQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqWE4kCRJDcOBJElqGA4kSVLDcCBJkhqGA0mS1DAcSJKkhuFAkiQ1DAeSJKlhOJAkSQ3DgSRJahgOJElSw3AgSZIahgNJktQwHEiSpIbhQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqWE4kCRJDcOBJElqGA4kSVJjouEgydlJNiXZneSyWX0nJ7kxyZ1JvpRkzUjfA5O8P8ntSbYkOWe+YyVJ0ngmPXPwPeAtwPtHG5McDlwJnAccBmwCPj5yyEbgaGAN8DTg9UlOnedYSZI0homGg6q6sqo+BfxgVtfzgc1VdUVV3U0XBo5Lckzf/xLgzVW1vapuAN4HvHSeYyVJ0himZc3BOuC6mRdVdQdwM7AuyaHAkaP9/fN1c41d5JolSVqWpiUcrAR2zGrbARzS9zGrf6ZvrrGNJGf0ax42bd26db+LliRpOZqWcLALWDWrbRWws+9jVv9M31xjG1V1SVWtr6r1q1ev3u+iJUlajqYlHGwGjpt5keRg4Ci6tQTbgdtG+/vnm+cau8g1S5K0LE36UsYVSQ4CDgAOSHJQkhXAJ4HHJtnQ978RuL6qbuyHfgg4N8mh/ULDVwKX9X1zjZUkSWOY9MzBucBdwBuAF/bPz62qrcAG4AJgO/Ak4LSRcW+iW2R4C3A18LaqugpgHmMlSdIYVkzyi1XVRrpLDffU9wVgj5cfVtVu4OX9Y6yxkiRpPNOy5kCSJE0Jw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqWE4kCRJDcOBJElqGA4kSVLDcCBJkhqGA0mS1DAcSJKkhuFAkiQ1DAeSJKlhOJAkSQ3DgSRJahgOJElSw3AgSZIahgNJktQwHEiSpIbhQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqSG4UCSJDUMB5IkqWE4kCRJDcOBJElqGA4kSVLDcCBJkhqGA0mS1DAcSJKkhuFAkiQ1DAeSJKlhOJAkSQ3DgSRJahgOJElSw3AgSZIahgNJktQwHEiSpMayCQdJDkvyySR3JLklya8NXZMkSUvRiqELWEAXAfcARwDHA59Lcl1VbR62LEmSlpZlMXOQ5GBgA3BeVe2qqmuATwMvGrYySZKWnlTV0DXstyRPAK6tqgePtL0WOKmqnj3SdgZwRv/y0cA/LVJJhwPbFum9F8NSqxeseRKWWr2w9GpeavWCNU/CYta7pqpWz3XQcjmtsBK4fVbbDuCQ0YaqugS4ZLGLSbKpqtYv9tdZKEutXrDmSVhq9cLSq3mp1QvWPAnTUO+yOK0A7AJWzWpbBewcoBZJkpa05RIObgJWJDl6pO04wMWIkiSNaVmEg6q6A7gSOD/JwUmeDPwS8CcDlbTopy4W2FKrF6x5EpZavbD0al5q9YI1T8Lg9S6LBYnQ3ecAeD/wDOAHwBuq6qPDViVJ0tKzbMKBJElaGMvitIIkSVo4hgNJktQwHCygpba/Q5Kzk2xKsjvJZUPXM5ckD0xyaf9vuzPJN5I8a+i65pLkw0luS3J7kpuSvGLomuYjydFJ7k7y4aFrmUuSv+pr3dU/FusGZwsqyWlJbuh/Ztyc5MSha9qTkX/XmcePk7xr6LrmkmRtks8n2Z5kS5J3J5na+/skOTbJF5PsSPLPSZ43VC2Gg4U1ur/D6cDFSdYNW9I+fQ94C91CzqVgBfAd4CTgIcC5wOVJ1g5Y03xcCKytqlXAc4C3JDlh4Jrm4yLgq0MXMYazq2pl/3j00MXMJckzgN8DXkZ3w7anAP8yaFF7MfLvuhL4OeAu4IqBy5qP9wDfB46k23PnJOCsQSvaiz60/BnwWeAwurv5fjjJo4aox3CwQJbi/g5VdWVVfYru6o6pV1V3VNXGqvq/VfWTqvos8C1gqn/RVtXmqto987J/HDVgSXNKchrwb8BfDl3LMva7wPlV9b/77+fvVtV3hy5qHjbQ/cL966ELmYd/D1xeVXdX1RbgKmBa/2A7Bvh3wNur6sdV9UXgWgb6HWI4WDiPAn5UVTeNtF3H9H4jLnlJjqD7d5/6m10leU+SO4EbgduAzw9c0l4lWQWcD5wzdC1jujDJtiTXJnnq0MXsS5IDgPXA6n76+NZ+yvtBQ9c2Dy8BPlRL41K3dwCnJXlwkocDz6ILCEtFgMcO8YUNBwtnXvs7aGEkORD4CPDBqrpx6HrmUlVn0X0vnEh3w67d+x4xqDcDl1bVrUMXMobfAh4JPJzuBjKfSTLNszNHAAcCL6D7njgeeALdqbKplWQN3dT8B4euZZ6+TPcH2u3ArcAm4FODVrR3/0Q3I/O6JAcmeSbdv/WD9z1scRgOFo77O0xIkgfQ3f3yHuDsgcuZt36q8Brg54Ezh65nT5IcDzwdePvQtYyjqv62qnZW1e6q+iDddOx/Hrqufbir/++7quq2qtoG/CHTXTN0U9zXVNW3hi5kLv3PiavowvjBdDsdHkq3zmPqVNW9wHOB/wJsAX4TuJwu1Eyc4WDhuL/DBCQJcCndX14b+v+hlpoVTO+ag6cCa4FvJ9kCvBbYkOTrQxZ1HxTdlOxUqqrtdD/0R6fml8I0/YtZOrMGhwGPAN7dh8YfAB9gigNYVV1fVSdV1UOr6hS62bC/G6IWw8ECmcL9HeaUZEWSg4ADgAOSHDTNl/n0LgaOBZ5dVXfNdfDQkjysv1xtZZIDkpwC/CrTu9DvErrgcnz/eC/wOeCUIYvalyQ/m+SUme/fJKfTrfyf9nPLHwBe3X+PHAr8d7qV6lMpyS/SnbZZClcp0M/GfAs4s/+++Fm69RLXD1vZ3iV5fP99/OAkr6W7yuKyIWoxHCyss4AH0Z03+hhwZlVN88zBuXTTm28AXtg/n9pznv35zlfR/dLaMnLN9ekDl7YvRXcK4VZgO/D7wG9U1acHrWovqurOqtoy86A7XXZ3VW0durZ9OJDuktytwDbg1cBzZy0OnkZvprtU9CbgBuDvgQsGrWjfXgJcWVVL6VTp84FT6b43/hm4ly6ETasX0S1Y/j5wMvCMkSudJsq9FSRJUsOZA0mS1DAcSJKkhuFAkiQ1DAeSJKlhOJAkSQ3DgSRJahgOJI0tyWVJ5nXDniRrk1SS9fv5NTcm+Yf9eQ9J82M4kCRJDcOBJElqGA4k7Zckpyb56yTbk/wwyZ8nOXYPhz4qyTVJ7k5yY78l7ej7PCbJ55LsTPL9JB9L8nMT+hiSRhgOJO2vg4F3AE+k29VxB/CZJD8z67i3Au+k2xvjL4A/S/JwgCRHAl8G/qF/n6cDK/tj/DklTdi078AnacpV1Z+Ovk7yMuB2ul/y14x0XVxVl/fHvIZup8cz6Tb7OhO4rqp+a+R9Xgz8EFjPQNvWSvdXJnJJ+yXJUUk+muTmJLcD/0r3s+URsw79m5knVfUT4G+Bx/RNJwBPGdlpcxfwnb7vqMX9BJJmc+ZA0v76LN2W1K8Cvgv8CPhHYPZphX15APA54LV76PvX/S1Q0ngMB5LusyQPBY4BzqqqL/Vt/4E9/2z5j8AX+2NCd9rhE33f14FfAW6pqnsXu25J++ZpBUn7YzuwDXhlkl9IchLwXrrZg9nOTPKCJI+mW8C4Bri477sIeAjw8SRPSvLIJE9PckmSQybwOSSNMBxIus/6tQP/FXg83ZUGFwHnAbv3cPgbgHOA64BTgedV1a39+3wPeDLwE+AqYHP/Xrv38l6SFlGqaugaJEnSFHHmQJIkNQwHkiSpYTiQJEkNw4EkSWoYDiRJUsNwIEmSGoYDSZLUMBxIkqTG/wO/eDnSn0kZfgAAAABJRU5ErkJggg==\n"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"* we don't assymetries in our distribution, all the numbers are equally represented\n\n#### getting 100 digits and plotting then to have an idea of the data we have, how the image looks like"},{"metadata":{"trusted":true},"cell_type":"code","source":"list_indx = [] #creating a list of the indexes for each digit\nfor i in range(10):\n    for nr in range(10):\n        indx = train_labels[train_labels==nr].index[i]\n        list_indx.append(indx) ","execution_count":8,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"fig, axs = plt.subplots(10, 10, sharex=True, sharey=True, figsize=(10,12))\naxs = axs.flatten() \nfor n, i in enumerate(list_indx): #n for each different plot and i for each different index stored in list_index\n    im = train.iloc[i]\n    im = im.values.reshape(-1,28,28,1) #reshaping it to 28x28 pixels\n    axs[n].imshow(im[0,:,:,0], cmap=plt.get_cmap('gray')) #making the background dark\n    axs[n].set_title(train_labels[i])\nplt.tight_layout()    ","execution_count":9,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 720x864 with 100 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"#### As can be seen, we have a varied of different digits for each digit\n##### most of them are easily identifiable, whereas some are not so, namely, 4s, 2s? 5s?"},{"metadata":{},"cell_type":"markdown","source":"## Separating a part of our training set to validate our models, normally we have 20% for evaluation and testing set, as the testing set was given, I am gonna separate 10% for the validation set"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.model_selection import train_test_split\n\ntrain, train_val, labels_train, labels_val = train_test_split(train, train_labels, test_size=0.1, random_state=42)","execution_count":10,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"***\n\n## creating a function that evaluates the prediction that has been made by different models and plot the confusion matrix"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.metrics import confusion_matrix, classification_report, accuracy_score\n\ndef report_and_confusion_matrix(label, prediction):\n    print(\"Model Report\")\n    print(classification_report(label, prediction))\n    score = accuracy_score(label, prediction)\n    print(\"Accuracy : \"+ str(score))\n    \n    ####################\n    fig, ax = plt.subplots(figsize=(8,8)) #setting the figure size and ax\n    mtx = confusion_matrix(label, prediction)\n    sns.heatmap(mtx, annot=True, fmt='d', linewidths=.5,  cbar=True, ax=ax) #create a heatmap with the values of our confusion matrix\n    plt.ylabel('true label')\n    plt.xlabel('predicted label')","execution_count":11,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"***\n### Models \n1. Logistic Regression"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.linear_model import LogisticRegression\nmodel_lgre = LogisticRegression(random_state=0)\nparam_grid = {'C': [0.014,0.012], 'multi_class': ['multinomial'],  \n              'penalty': ['l1'],'solver': ['saga'], 'tol': [0.1] }\nGridCV_LR = GridSearchCV(model_lgre, param_grid, verbose=1, cv=5)\nGridCV_LR.fit(train,labels_train)\nscore_grid_LR = GridCV_LR.best_score_","execution_count":12,"outputs":[{"output_type":"stream","text":"Fitting 5 folds for each of 2 candidates, totalling 10 fits\n","name":"stdout"},{"output_type":"stream","text":"[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n[Parallel(n_jobs=1)]: Done  10 out of  10 | elapsed:  1.4min finished\n","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(score_grid_LR)","execution_count":13,"outputs":[{"output_type":"stream","text":"0.8787566137566137\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"##using the logist regression parameters found in our grid search, let's predict using our validation set and see how well the model is doing\npredict_lgr = GridCV_LR.predict(train_val)\nreport_and_confusion_matrix(labels_val, predict_lgr)","execution_count":14,"outputs":[{"output_type":"stream","text":"Model Report\n              precision    recall  f1-score   support\n\n           0       0.93      0.95      0.94       408\n           1       0.91      0.97      0.94       471\n           2       0.89      0.86      0.87       420\n           3       0.89      0.83      0.85       506\n           4       0.83      0.91      0.87       397\n           5       0.83      0.83      0.83       339\n           6       0.90      0.90      0.90       402\n           7       0.89      0.88      0.88       438\n           8       0.87      0.83      0.85       403\n           9       0.85      0.84      0.85       416\n\n   micro avg       0.88      0.88      0.88      4200\n   macro avg       0.88      0.88      0.88      4200\nweighted avg       0.88      0.88      0.88      4200\n\nAccuracy : 0.8809523809523809\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"#### As can be seen, the model is generalizig fairly well, the confusion matrix is showing that some digits have like 1s, are easier to predict and 5s not quite so.(this is due to the shapes of some digits)"},{"metadata":{},"cell_type":"markdown","source":"2. K NeighborsClassifier"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.neighbors import KNeighborsClassifier\nknn_clf = KNeighborsClassifier(weights='distance', n_neighbors=4)\nknn_clf.fit(train, labels_train)","execution_count":15,"outputs":[{"output_type":"execute_result","execution_count":15,"data":{"text/plain":"KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',\n           metric_params=None, n_jobs=None, n_neighbors=4, p=2,\n           weights='distance')"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"predict_knn = knn_clf.predict(train_val)\nreport_and_confusion_matrix(labels_val, predict_knn)","execution_count":16,"outputs":[{"output_type":"stream","text":"Model Report\n              precision    recall  f1-score   support\n\n           0       0.98      1.00      0.99       408\n           1       0.95      1.00      0.97       471\n           2       0.99      0.96      0.97       420\n           3       0.97      0.96      0.97       506\n           4       0.99      0.97      0.98       397\n           5       0.96      0.97      0.96       339\n           6       0.97      0.99      0.98       402\n           7       0.96      0.97      0.96       438\n           8       0.98      0.92      0.95       403\n           9       0.95      0.97      0.96       416\n\n   micro avg       0.97      0.97      0.97      4200\n   macro avg       0.97      0.97      0.97      4200\nweighted avg       0.97      0.97      0.97      4200\n\nAccuracy : 0.9702380952380952\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"### A significant improvement compared to logistic regression...\n#### Now, random forest? \n\n3. The random forest model"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.ensemble import RandomForestClassifier\n\nrand_forest_clf = RandomForestClassifier(random_state=42)\nparam_grid = {'max_depth': [15], 'max_features': [100],  \n              'min_samples_split': [5],'n_estimators' : [50] }\nGridCV_rd_clf = GridSearchCV(rand_forest_clf, param_grid, verbose=1, cv=5)\nGridCV_rd_clf.fit(train, labels_train)\nscore_grid_rd = GridCV_rd_clf.best_score_","execution_count":17,"outputs":[{"output_type":"stream","text":"Fitting 5 folds for each of 1 candidates, totalling 5 fits\n","name":"stdout"},{"output_type":"stream","text":"[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:  3.0min finished\n","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(score_grid_rd)","execution_count":18,"outputs":[{"output_type":"stream","text":"0.9566402116402116\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"predict_rand_forest = GridCV_rd_clf.predict(train_val)\nreport_and_confusion_matrix(labels_val, predict_rand_forest)","execution_count":19,"outputs":[{"output_type":"stream","text":"Model Report\n              precision    recall  f1-score   support\n\n           0       0.98      0.99      0.98       408\n           1       0.97      0.99      0.98       471\n           2       0.95      0.96      0.96       420\n           3       0.97      0.94      0.95       506\n           4       0.95      0.96      0.96       397\n           5       0.96      0.96      0.96       339\n           6       0.98      0.98      0.98       402\n           7       0.96      0.95      0.96       438\n           8       0.94      0.94      0.94       403\n           9       0.92      0.93      0.93       416\n\n   micro avg       0.96      0.96      0.96      4200\n   macro avg       0.96      0.96      0.96      4200\nweighted avg       0.96      0.96      0.96      4200\n\nAccuracy : 0.9597619047619048\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"* The random forest model is generalizing a bit better than the KNN"},{"metadata":{},"cell_type":"markdown","source":"4. Multi Layer Perceptron"},{"metadata":{"trusted":true},"cell_type":"code","source":"from sklearn.neural_network import MLPClassifier\n\nmlp_clf = MLPClassifier(activation = \"logistic\", hidden_layer_sizes=(200,), random_state=42,batch_size=2000)\nparam_grid = { 'max_iter': [600], 'alpha': [1e-4], \n               'solver': ['sgd'], 'learning_rate_init': [0.05,0.06],'tol': [1e-4] }\n    \nGridCV_MLP = GridSearchCV(mlp_clf, param_grid, verbose=1, cv=3)\nGridCV_MLP.fit(train,labels_train)\nscore_grid_MLP = GridCV_MLP.best_score_","execution_count":20,"outputs":[{"output_type":"stream","text":"Fitting 3 folds for each of 2 candidates, totalling 6 fits\n","name":"stdout"},{"output_type":"stream","text":"[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n[Parallel(n_jobs=1)]: Done   6 out of   6 | elapsed: 45.6min finished\n","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"print(score_grid_MLP)","execution_count":21,"outputs":[{"output_type":"stream","text":"0.9679629629629629\n","name":"stdout"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"predict_MLP = GridCV_MLP.predict(train_val)\nreport_and_confusion_matrix(labels_val, predict_MLP)","execution_count":22,"outputs":[{"output_type":"stream","text":"Model Report\n              precision    recall  f1-score   support\n\n           0       0.98      0.99      0.98       408\n           1       0.98      0.99      0.99       471\n           2       0.99      0.97      0.98       420\n           3       0.97      0.96      0.97       506\n           4       0.98      0.97      0.97       397\n           5       0.97      0.97      0.97       339\n           6       0.98      0.98      0.98       402\n           7       0.96      0.97      0.97       438\n           8       0.97      0.97      0.97       403\n           9       0.96      0.97      0.96       416\n\n   micro avg       0.97      0.97      0.97      4200\n   macro avg       0.97      0.97      0.97      4200\nweighted avg       0.97      0.97      0.97      4200\n\nAccuracy : 0.9735714285714285\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"predict_test = GridCV_MLP.predict(test)","execution_count":23,"outputs":[]},{"metadata":{},"cell_type":"markdown","source":"* Our multi layer perceptron gave us the best result so far..."},{"metadata":{"trusted":true},"cell_type":"code","source":"sub = pd.read_csv('../input/sample_submission.csv')\nsub['Label'] = predict_test.astype('int64')\nsub.to_csv(\"mlp.csv\", index=False)","execution_count":24,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"## loading relevant models for this part of the notebook\nfrom keras.models import Sequential, Model\nfrom keras.optimizers import Adam, SGD\nfrom keras.layers import Dense, Activation, Dropout, Input, concatenate\nfrom keras.utils.np_utils import to_categorical\nfrom keras.wrappers.scikit_learn import KerasRegressor\nfrom keras.callbacks import EarlyStopping","execution_count":25,"outputs":[{"output_type":"stream","text":"Using TensorFlow backend.\n","name":"stderr"}]},{"metadata":{"trusted":true},"cell_type":"code","source":"def build_model(input_shape, n_hidden=1, n_neurons=30,optimizer = SGD(3e-3)):\n    model = Sequential()\n    options = {\"input_shape\": input_shape}\n    for layer in range(n_hidden):\n        model.add(Dense(n_neurons, activation=\"elu\", kernel_initializer=\"he_normal\", **options))\n        model.add(Dropout(0.25))\n        options = {}\n    model.add(Dense(10, activation='softmax',**options))\n    model.compile(loss=\"categorical_crossentropy\", optimizer=optimizer,metrics=['accuracy'])\n    return model","execution_count":37,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"y_train = to_categorical(labels_train, 10)\ny_val = to_categorical(labels_val, 10)","execution_count":28,"outputs":[]},{"metadata":{"trusted":true},"cell_type":"code","source":"model = build_model(input_shape=train.shape[1:], n_hidden=4, n_neurons=60, optimizer='Adam')\nhistory = model.fit(x=train, y=y_train, epochs=60, validation_data=(train_val, y_val))","execution_count":42,"outputs":[{"output_type":"stream","text":"Train on 37800 samples, validate on 4200 samples\nEpoch 1/60\n37800/37800 [==============================] - 6s 146us/step - loss: 0.6595 - acc: 0.7904 - val_loss: 0.2788 - val_acc: 0.9131\nEpoch 2/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.3795 - acc: 0.8872 - val_loss: 0.2268 - val_acc: 0.9293\nEpoch 3/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.3157 - acc: 0.9075 - val_loss: 0.1889 - val_acc: 0.9410\nEpoch 4/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.2812 - acc: 0.9148 - val_loss: 0.1857 - val_acc: 0.9438\nEpoch 5/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.2534 - acc: 0.9262 - val_loss: 0.1750 - val_acc: 0.9471\nEpoch 6/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.2293 - acc: 0.9322 - val_loss: 0.1633 - val_acc: 0.9488\nEpoch 7/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.2131 - acc: 0.9374 - val_loss: 0.1531 - val_acc: 0.9555\nEpoch 8/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.2019 - acc: 0.9403 - val_loss: 0.1496 - val_acc: 0.9543\nEpoch 9/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1890 - acc: 0.9447 - val_loss: 0.1300 - val_acc: 0.9607\nEpoch 10/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1770 - acc: 0.9488 - val_loss: 0.1328 - val_acc: 0.9598\nEpoch 11/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1741 - acc: 0.9489 - val_loss: 0.1382 - val_acc: 0.9588\nEpoch 12/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1634 - acc: 0.9518 - val_loss: 0.1247 - val_acc: 0.9617\nEpoch 13/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1607 - acc: 0.9531 - val_loss: 0.1297 - val_acc: 0.9633\nEpoch 14/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1508 - acc: 0.9554 - val_loss: 0.1368 - val_acc: 0.9595\nEpoch 15/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1458 - acc: 0.9567 - val_loss: 0.1213 - val_acc: 0.9671\nEpoch 16/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1392 - acc: 0.9593 - val_loss: 0.1221 - val_acc: 0.9655\nEpoch 17/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1386 - acc: 0.9585 - val_loss: 0.1134 - val_acc: 0.9667\nEpoch 18/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1351 - acc: 0.9609 - val_loss: 0.1263 - val_acc: 0.9669\nEpoch 19/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1317 - acc: 0.9611 - val_loss: 0.1087 - val_acc: 0.9679\nEpoch 20/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1247 - acc: 0.9627 - val_loss: 0.1041 - val_acc: 0.9698\nEpoch 21/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1243 - acc: 0.9633 - val_loss: 0.1067 - val_acc: 0.9688\nEpoch 22/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.1251 - acc: 0.9636 - val_loss: 0.1141 - val_acc: 0.9671\nEpoch 23/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1162 - acc: 0.9656 - val_loss: 0.1149 - val_acc: 0.9681\nEpoch 24/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1154 - acc: 0.9665 - val_loss: 0.1066 - val_acc: 0.9698\nEpoch 25/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1130 - acc: 0.9667 - val_loss: 0.1151 - val_acc: 0.9695\nEpoch 26/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1113 - acc: 0.9676 - val_loss: 0.1049 - val_acc: 0.9712\nEpoch 27/60\n37800/37800 [==============================] - 4s 115us/step - loss: 0.1069 - acc: 0.9681 - val_loss: 0.1169 - val_acc: 0.9686\nEpoch 28/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.1093 - acc: 0.9682 - val_loss: 0.1078 - val_acc: 0.9712\nEpoch 29/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.1050 - acc: 0.9692 - val_loss: 0.1093 - val_acc: 0.9702\nEpoch 30/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.1044 - acc: 0.9692 - val_loss: 0.1059 - val_acc: 0.9702\nEpoch 31/60\n37800/37800 [==============================] - 5s 128us/step - loss: 0.1047 - acc: 0.9695 - val_loss: 0.1092 - val_acc: 0.9700\nEpoch 32/60\n37800/37800 [==============================] - 5s 128us/step - loss: 0.0996 - acc: 0.9708 - val_loss: 0.1085 - val_acc: 0.9705\nEpoch 33/60\n37800/37800 [==============================] - 4s 115us/step - loss: 0.0980 - acc: 0.9714 - val_loss: 0.1106 - val_acc: 0.9707\nEpoch 34/60\n37800/37800 [==============================] - 5s 121us/step - loss: 0.1011 - acc: 0.9713 - val_loss: 0.1145 - val_acc: 0.9705\nEpoch 35/60\n37800/37800 [==============================] - 4s 115us/step - loss: 0.0980 - acc: 0.9708 - val_loss: 0.1102 - val_acc: 0.9721\nEpoch 36/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0938 - acc: 0.9726 - val_loss: 0.1072 - val_acc: 0.9719\nEpoch 37/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0952 - acc: 0.9719 - val_loss: 0.0995 - val_acc: 0.9714\nEpoch 38/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0941 - acc: 0.9727 - val_loss: 0.1101 - val_acc: 0.9700\nEpoch 39/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0937 - acc: 0.9726 - val_loss: 0.1102 - val_acc: 0.9707\nEpoch 40/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0938 - acc: 0.9722 - val_loss: 0.1078 - val_acc: 0.9717\nEpoch 41/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0925 - acc: 0.9734 - val_loss: 0.1138 - val_acc: 0.9707\nEpoch 42/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0863 - acc: 0.9749 - val_loss: 0.1096 - val_acc: 0.9719\nEpoch 43/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0891 - acc: 0.9738 - val_loss: 0.1023 - val_acc: 0.9731\nEpoch 44/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0847 - acc: 0.9741 - val_loss: 0.1060 - val_acc: 0.9721\nEpoch 45/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0877 - acc: 0.9743 - val_loss: 0.1049 - val_acc: 0.9726\nEpoch 46/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0864 - acc: 0.9749 - val_loss: 0.1032 - val_acc: 0.9724\nEpoch 47/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0877 - acc: 0.9744 - val_loss: 0.1000 - val_acc: 0.9729\nEpoch 48/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0834 - acc: 0.9752 - val_loss: 0.1029 - val_acc: 0.9726\nEpoch 49/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0831 - acc: 0.9748 - val_loss: 0.1033 - val_acc: 0.9724\nEpoch 50/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0797 - acc: 0.9762 - val_loss: 0.1106 - val_acc: 0.9719\nEpoch 51/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.0810 - acc: 0.9767 - val_loss: 0.1091 - val_acc: 0.9729\nEpoch 52/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0816 - acc: 0.9751 - val_loss: 0.1061 - val_acc: 0.9717\nEpoch 53/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0807 - acc: 0.9767 - val_loss: 0.1077 - val_acc: 0.9717\nEpoch 54/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0830 - acc: 0.9748 - val_loss: 0.0962 - val_acc: 0.9733\nEpoch 55/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0797 - acc: 0.9762 - val_loss: 0.0894 - val_acc: 0.9750\nEpoch 56/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.0819 - acc: 0.9753 - val_loss: 0.0974 - val_acc: 0.9743\nEpoch 57/60\n37800/37800 [==============================] - 4s 114us/step - loss: 0.0769 - acc: 0.9785 - val_loss: 0.1109 - val_acc: 0.9719\nEpoch 58/60\n37800/37800 [==============================] - 4s 113us/step - loss: 0.0787 - acc: 0.9767 - val_loss: 0.1034 - val_acc: 0.9731\nEpoch 59/60\n","name":"stdout"},{"output_type":"stream","text":"37800/37800 [==============================] - 4s 112us/step - loss: 0.0770 - acc: 0.9774 - val_loss: 0.0989 - val_acc: 0.9731\nEpoch 60/60\n37800/37800 [==============================] - 4s 112us/step - loss: 0.0753 - acc: 0.9773 - val_loss: 0.0999 - val_acc: 0.9755\n","name":"stdout"}]},{"metadata":{},"cell_type":"markdown","source":"#### plotting the accuracy, loss in both training set and validation set"},{"metadata":{"trusted":true},"cell_type":"code","source":"pd.DataFrame(history.history).plot(figsize=(8, 5))\nplt.grid(True)\nplt.gca().set_ylim(0, 1) # set the vertical range to [0-1]","execution_count":43,"outputs":[{"output_type":"execute_result","execution_count":43,"data":{"text/plain":"(0, 1)"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x360 with 1 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{},"cell_type":"markdown","source":"** it can be seen that the model is overfitting, the validation loss is telling us that, in order to get a better score I would need to use some regularization, or early stopping**"},{"metadata":{"trusted":true},"cell_type":"code","source":"prediction_keras_nn = model.predict_classes(train_val)\nreport_and_confusion_matrix(labels_val, prediction_keras_nn)","execution_count":44,"outputs":[{"output_type":"stream","text":"Model Report\n              precision    recall  f1-score   support\n\n           0       0.98      0.98      0.98       408\n           1       0.99      0.99      0.99       471\n           2       0.98      0.97      0.98       420\n           3       0.97      0.97      0.97       506\n           4       0.97      0.97      0.97       397\n           5       0.96      0.98      0.97       339\n           6       0.98      0.99      0.98       402\n           7       0.97      0.97      0.97       438\n           8       0.96      0.96      0.96       403\n           9       0.98      0.97      0.97       416\n\n   micro avg       0.98      0.98      0.98      4200\n   macro avg       0.98      0.98      0.98      4200\nweighted avg       0.98      0.98      0.98      4200\n\nAccuracy : 0.9754761904761905\n","name":"stdout"},{"output_type":"display_data","data":{"text/plain":"<Figure size 576x576 with 2 Axes>","image/png":"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\n"},"metadata":{}}]},{"metadata":{"trusted":true},"cell_type":"code","source":"prediction_keras_nn_test_sub = model.predict_classes(test)\nsub['Label'] = prediction_keras_nn_test_sub.astype('int64')\nsub.to_csv(\"Keras_NN.csv\", index=False)","execution_count":45,"outputs":[]}],"metadata":{"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"name":"python","version":"3.6.4","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat":4,"nbformat_minor":1}