custom_points_model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ZObnILgST2-y"
   },
   "source": [
    "# Summary : File for training the  posenet coordinates and generating the model for classification."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "from sklearn import preprocessing\n",
    "from sklearn.model_selection import train_test_split\n",
    "# first neural network with keras make predictions\n",
    "from numpy import loadtxt\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense, Dropout\n",
    "from keras.layers import Dense\n",
    "from keras.wrappers.scikit_learn import KerasClassifier\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import KFold\n",
    "from keras.optimizers import SGD, Adam, Adadelta, RMSprop\n",
    "\n",
    "from sklearn.metrics import confusion_matrix\n",
    "from sklearn.metrics import classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "id": "LwmPkeOST9jq"
   },
   "outputs": [],
   "source": [
    "#Provide link to the dataset from google drive. \n",
    "df = pd.read_csv(\"dataset.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 317
    },
    "id": "z3WhRFBuUBjS",
    "outputId": "b569b283-7ecf-44d3-93eb-66e14d315491"
   },
   "outputs": [
    {
     "data": {
      "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>Unnamed: 0</th>\n",
       "      <th>nose_x</th>\n",
       "      <th>nose_y</th>\n",
       "      <th>left_eye_x</th>\n",
       "      <th>left_eye_y</th>\n",
       "      <th>right_eye_x</th>\n",
       "      <th>right_eye_y</th>\n",
       "      <th>left_ear_x</th>\n",
       "      <th>left_ear_y</th>\n",
       "      <th>right_ear_x</th>\n",
       "      <th>...</th>\n",
       "      <th>right_hip_y</th>\n",
       "      <th>left_knee_x</th>\n",
       "      <th>left_knee_y</th>\n",
       "      <th>right_knee_x</th>\n",
       "      <th>right_knee_y</th>\n",
       "      <th>left_ankle_x</th>\n",
       "      <th>left_ankle_y</th>\n",
       "      <th>right_ankle_x</th>\n",
       "      <th>right_ankle_y</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0.649173</td>\n",
       "      <td>0.523219</td>\n",
       "      <td>0.657339</td>\n",
       "      <td>0.503735</td>\n",
       "      <td>0.655636</td>\n",
       "      <td>0.504516</td>\n",
       "      <td>0.630299</td>\n",
       "      <td>0.455352</td>\n",
       "      <td>0.636445</td>\n",
       "      <td>...</td>\n",
       "      <td>0.666943</td>\n",
       "      <td>0.540174</td>\n",
       "      <td>0.773534</td>\n",
       "      <td>0.539856</td>\n",
       "      <td>0.762104</td>\n",
       "      <td>0.699938</td>\n",
       "      <td>0.878527</td>\n",
       "      <td>0.673206</td>\n",
       "      <td>0.852668</td>\n",
       "      <td>downdog</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0.655792</td>\n",
       "      <td>0.575658</td>\n",
       "      <td>0.660437</td>\n",
       "      <td>0.590248</td>\n",
       "      <td>0.655015</td>\n",
       "      <td>0.590673</td>\n",
       "      <td>0.636295</td>\n",
       "      <td>0.615927</td>\n",
       "      <td>0.623701</td>\n",
       "      <td>...</td>\n",
       "      <td>0.426795</td>\n",
       "      <td>0.543175</td>\n",
       "      <td>0.299537</td>\n",
       "      <td>0.544742</td>\n",
       "      <td>0.302838</td>\n",
       "      <td>0.675126</td>\n",
       "      <td>0.180818</td>\n",
       "      <td>0.675930</td>\n",
       "      <td>0.177933</td>\n",
       "      <td>downdog</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0.659473</td>\n",
       "      <td>0.599582</td>\n",
       "      <td>0.660218</td>\n",
       "      <td>0.620654</td>\n",
       "      <td>0.662387</td>\n",
       "      <td>0.619791</td>\n",
       "      <td>0.615932</td>\n",
       "      <td>0.658006</td>\n",
       "      <td>0.614723</td>\n",
       "      <td>...</td>\n",
       "      <td>0.379610</td>\n",
       "      <td>0.537431</td>\n",
       "      <td>0.221986</td>\n",
       "      <td>0.542609</td>\n",
       "      <td>0.222733</td>\n",
       "      <td>0.732093</td>\n",
       "      <td>0.105802</td>\n",
       "      <td>0.746955</td>\n",
       "      <td>0.084199</td>\n",
       "      <td>downdog</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0.671158</td>\n",
       "      <td>0.542812</td>\n",
       "      <td>0.679293</td>\n",
       "      <td>0.560280</td>\n",
       "      <td>0.676892</td>\n",
       "      <td>0.559193</td>\n",
       "      <td>0.651990</td>\n",
       "      <td>0.601257</td>\n",
       "      <td>0.644605</td>\n",
       "      <td>...</td>\n",
       "      <td>0.411304</td>\n",
       "      <td>0.529392</td>\n",
       "      <td>0.294301</td>\n",
       "      <td>0.532135</td>\n",
       "      <td>0.295052</td>\n",
       "      <td>0.703284</td>\n",
       "      <td>0.193187</td>\n",
       "      <td>0.710546</td>\n",
       "      <td>0.181630</td>\n",
       "      <td>downdog</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>0.607264</td>\n",
       "      <td>0.357824</td>\n",
       "      <td>0.612975</td>\n",
       "      <td>0.337642</td>\n",
       "      <td>0.614474</td>\n",
       "      <td>0.339381</td>\n",
       "      <td>0.570222</td>\n",
       "      <td>0.307927</td>\n",
       "      <td>0.576377</td>\n",
       "      <td>...</td>\n",
       "      <td>0.563056</td>\n",
       "      <td>0.485910</td>\n",
       "      <td>0.705451</td>\n",
       "      <td>0.492973</td>\n",
       "      <td>0.701507</td>\n",
       "      <td>0.673677</td>\n",
       "      <td>0.831081</td>\n",
       "      <td>0.666935</td>\n",
       "      <td>0.816390</td>\n",
       "      <td>downdog</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 36 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0    nose_x    nose_y  left_eye_x  left_eye_y  right_eye_x  \\\n",
       "0           0  0.649173  0.523219    0.657339    0.503735     0.655636   \n",
       "1           1  0.655792  0.575658    0.660437    0.590248     0.655015   \n",
       "2           2  0.659473  0.599582    0.660218    0.620654     0.662387   \n",
       "3           3  0.671158  0.542812    0.679293    0.560280     0.676892   \n",
       "4           4  0.607264  0.357824    0.612975    0.337642     0.614474   \n",
       "\n",
       "   right_eye_y  left_ear_x  left_ear_y  right_ear_x  ...  right_hip_y  \\\n",
       "0     0.504516    0.630299    0.455352     0.636445  ...     0.666943   \n",
       "1     0.590673    0.636295    0.615927     0.623701  ...     0.426795   \n",
       "2     0.619791    0.615932    0.658006     0.614723  ...     0.379610   \n",
       "3     0.559193    0.651990    0.601257     0.644605  ...     0.411304   \n",
       "4     0.339381    0.570222    0.307927     0.576377  ...     0.563056   \n",
       "\n",
       "   left_knee_x  left_knee_y  right_knee_x  right_knee_y  left_ankle_x  \\\n",
       "0     0.540174     0.773534      0.539856      0.762104      0.699938   \n",
       "1     0.543175     0.299537      0.544742      0.302838      0.675126   \n",
       "2     0.537431     0.221986      0.542609      0.222733      0.732093   \n",
       "3     0.529392     0.294301      0.532135      0.295052      0.703284   \n",
       "4     0.485910     0.705451      0.492973      0.701507      0.673677   \n",
       "\n",
       "   left_ankle_y  right_ankle_x  right_ankle_y   target  \n",
       "0      0.878527       0.673206       0.852668  downdog  \n",
       "1      0.180818       0.675930       0.177933  downdog  \n",
       "2      0.105802       0.746955       0.084199  downdog  \n",
       "3      0.193187       0.710546       0.181630  downdog  \n",
       "4      0.831081       0.666935       0.816390  downdog  \n",
       "\n",
       "[5 rows x 36 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "id": "0Zs3yYgsWRHU"
   },
   "outputs": [],
   "source": [
    "df.index = df['Unnamed: 0']\n",
    "df.drop('Unnamed: 0', axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "9U7_PR6YUKjp",
    "outputId": "ec4c1532-5b54-4a1b-c14b-649c19c030e6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Drop :  nose_x\n",
      "Drop :  nose_y\n",
      "Drop :  left_eye_x\n",
      "Drop :  left_eye_y\n",
      "Drop :  right_eye_x\n",
      "Drop :  right_eye_y\n",
      "Drop :  left_ear_x\n",
      "Drop :  left_ear_y\n",
      "Drop :  right_ear_x\n",
      "Drop :  right_ear_y\n"
     ]
    }
   ],
   "source": [
    "#Cleaning the data => Removing columns with nose, ear and eye\n",
    "points_to_remove = ['nose','ear','eye']\n",
    "for c in df.columns:\n",
    "       for point in points_to_remove:\n",
    "         if(point in c):\n",
    "           print(\"Drop : \",c)\n",
    "           df.drop(c, axis=1, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "o2xhD7Dub3kM",
    "outputId": "6acf41e4-aaff-4013-a039-54e3383cc5fb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(470, 25)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape\n",
    "#25 columns in the data frame => taking 24 co ordinates to detect a pose"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "rn4Q2lDIWOtR",
    "outputId": "d850af37-cff9-497b-d066-b133b4d19ca0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['left_shoulder_x', 'left_shoulder_y', 'right_shoulder_x',\n",
       "       'right_shoulder_y', 'left_elbow_x', 'left_elbow_y', 'right_elbow_x',\n",
       "       'right_elbow_y', 'left_wrist_x', 'left_wrist_y', 'right_wrist_x',\n",
       "       'right_wrist_y', 'left_hip_x', 'left_hip_y', 'right_hip_x',\n",
       "       'right_hip_y', 'left_knee_x', 'left_knee_y', 'right_knee_x',\n",
       "       'right_knee_y', 'left_ankle_x', 'left_ankle_y', 'right_ankle_x',\n",
       "       'right_ankle_y', 'target'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 295
    },
    "id": "P-udjdtXWWtm",
    "outputId": "df22ee5b-aa35-48c7-a20d-1b33c1bc663e"
   },
   "outputs": [
    {
     "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": [
    "#Showing the number of images present for each yoga pose \n",
    "df.target.value_counts().plot(kind='barh')\n",
    "plt.title(\"Distribution of the data \")\n",
    "plt.xlabel(\"Count of the Poses\")\n",
    "plt.ylabel(\"Target Labels\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "057J237bWYyc"
   },
   "outputs": [],
   "source": [
    "#Separating input(X=>all the coordinates of joints) and output(Y=>pose) data \n",
    "X = df.drop(columns=['target'])\n",
    "y_temp = df['target']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "tUNTzLwqrHnc"
   },
   "outputs": [],
   "source": [
    "#Encoding the target variables with labels for each class \n",
    "le = preprocessing.LabelEncoder()\n",
    "y = le.fit_transform(y_temp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "vJzSRP6Plb7N",
    "outputId": "c31a0939-055f-4e1f-e499-ac6a6010e8a6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 downdog\n",
      "1 goddess\n",
      "2 plank\n",
      "3 tree\n",
      "4 warrior2\n"
     ]
    }
   ],
   "source": [
    "#showing labels assigned to each pose \n",
    "poses = ['downdog', 'goddess', 'plank', 'tree', 'warrior2']\n",
    "encoded_data, mapping_index = pd.Series(poses).factorize()\n",
    "for data in encoded_data:\n",
    "    print(data, mapping_index[data])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 971
    },
    "id": "YTXQDu-4jXSb",
    "outputId": "31ebaa20-1270-4998-e06d-4be846f472ba"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1368x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#plotting the correlation matrix \n",
    "f = plt.figure(figsize=(19, 15))\n",
    "plt.matshow(df.corr(), fignum=f.number)\n",
    "search = '_x'\n",
    "\n",
    "plt.xticks(range(df.select_dtypes(['number']).shape[1]), df.select_dtypes(['number']).columns, fontsize=14, rotation=45)\n",
    "plt.yticks(range(df.select_dtypes(['number']).shape[1]), df.select_dtypes(['number']).columns, fontsize=14)\n",
    "cb = plt.colorbar()\n",
    "cb.ax.tick_params(labelsize=14)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "eavRug9pj8Kp"
   },
   "source": [
    "#Training Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "J0lptUiUdZVZ"
   },
   "outputs": [],
   "source": [
    "#splitting the train and test datas \n",
    "X_dev, X_test, y_dev, y_test = train_test_split( X, y, test_size=0.20, random_state=42)\n",
    "#Splitting the data into validation and train test\n",
    "X_train, X_val, y_train, y_val = train_test_split( X_dev, y_dev, test_size=0.20, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "GfsgXCsZg_f7",
    "outputId": "d23a27ed-bab6-4938-e5a1-6a820ad94b69"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shape of the dataset\n",
      "Train  (300, 24)\n",
      "Test  (94, 24)\n",
      "Validation  (76, 24)\n"
     ]
    }
   ],
   "source": [
    "print(\"Shape of the dataset\")\n",
    "print(\"Train \", X_train.shape)\n",
    "print(\"Test \", X_test.shape)\n",
    "print(\"Validation \", X_val.shape)\n",
    "input_shape = X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "GkMchbPtWt0K",
    "outputId": "3945480c-7b21-4487-a84a-8752e7547458"
   },
   "outputs": [],
   "source": [
    "#Building custom model \n",
    "model = Sequential()\n",
    "model.add(Dense(32,  activation = \"relu\"))\n",
    "model.add(Dense(128, activation = \"relu\"))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Dense(32, activation = \"relu\"))\n",
    "model.add(Dense(128, activation = \"relu\"))\n",
    "model.add(Dense(64, activation = \"relu\"))\n",
    "model.add(Dropout(0.1))\n",
    "model.add(Dense(32, activation = \"softmax\"))\n",
    "model.compile(Adam(lr = 0.01), \"sparse_categorical_crossentropy\", metrics = [\"accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "-4jY-BE-hhBX",
    "outputId": "a9ce2d0c-036e-4848-bbce-d40077d1c2d0"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_2\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_12 (Dense)             (300, 32)                 800       \n",
      "_________________________________________________________________\n",
      "dense_13 (Dense)             (300, 128)                4224      \n",
      "_________________________________________________________________\n",
      "dropout_4 (Dropout)          (300, 128)                0         \n",
      "_________________________________________________________________\n",
      "dense_14 (Dense)             (300, 32)                 4128      \n",
      "_________________________________________________________________\n",
      "dense_15 (Dense)             (300, 128)                4224      \n",
      "_________________________________________________________________\n",
      "dense_16 (Dense)             (300, 64)                 8256      \n",
      "_________________________________________________________________\n",
      "dropout_5 (Dropout)          (300, 64)                 0         \n",
      "_________________________________________________________________\n",
      "dense_17 (Dense)             (300, 32)                 2080      \n",
      "=================================================================\n",
      "Total params: 23,712\n",
      "Trainable params: 23,712\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.build(input_shape)\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "a5JgJHBjW0SZ",
    "outputId": "c6899034-3222-4be1-baf0-2cdf371de1a6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/60\n",
      "WARNING:tensorflow:Layer dense_12 is casting an input tensor from dtype float64 to the layer's dtype of float32, which is new behavior in TensorFlow 2.  The layer has dtype float32 because its dtype defaults to floatx.\n",
      "\n",
      "If you intended to run this layer in float32, you can safely ignore this warning. If in doubt, this warning is likely only an issue if you are porting a TensorFlow 1.X model to TensorFlow 2.\n",
      "\n",
      "To change all layers to have dtype float64 by default, call `tf.keras.backend.set_floatx('float64')`. To change just this layer, pass dtype='float64' to the layer constructor. If you are the author of this layer, you can disable autocasting by passing autocast=False to the base Layer constructor.\n",
      "\n",
      "10/10 [==============================] - 0s 14ms/step - loss: 2.3488 - accuracy: 0.2033 - val_loss: 1.6653 - val_accuracy: 0.4868\n",
      "Epoch 2/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.6484 - accuracy: 0.2367 - val_loss: 1.5752 - val_accuracy: 0.3947\n",
      "Epoch 3/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.4338 - accuracy: 0.3533 - val_loss: 1.2815 - val_accuracy: 0.3421\n",
      "Epoch 4/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.3119 - accuracy: 0.3467 - val_loss: 1.1339 - val_accuracy: 0.3947\n",
      "Epoch 5/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.1566 - accuracy: 0.3900 - val_loss: 1.0775 - val_accuracy: 0.3289\n",
      "Epoch 6/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.0785 - accuracy: 0.4167 - val_loss: 1.0429 - val_accuracy: 0.5263\n",
      "Epoch 7/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.0226 - accuracy: 0.4233 - val_loss: 1.0368 - val_accuracy: 0.5395\n",
      "Epoch 8/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.9763 - accuracy: 0.5500 - val_loss: 0.9344 - val_accuracy: 0.5395\n",
      "Epoch 9/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.9566 - accuracy: 0.5100 - val_loss: 0.9158 - val_accuracy: 0.4868\n",
      "Epoch 10/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.0319 - accuracy: 0.5433 - val_loss: 0.9441 - val_accuracy: 0.5921\n",
      "Epoch 11/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.9627 - accuracy: 0.5800 - val_loss: 0.8499 - val_accuracy: 0.6447\n",
      "Epoch 12/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.8065 - accuracy: 0.6100 - val_loss: 0.6804 - val_accuracy: 0.7368\n",
      "Epoch 13/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.7307 - accuracy: 0.6767 - val_loss: 0.7266 - val_accuracy: 0.6053\n",
      "Epoch 14/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6930 - accuracy: 0.7033 - val_loss: 0.7779 - val_accuracy: 0.6842\n",
      "Epoch 15/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6463 - accuracy: 0.7300 - val_loss: 0.5914 - val_accuracy: 0.7632\n",
      "Epoch 16/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.5707 - accuracy: 0.7667 - val_loss: 0.5409 - val_accuracy: 0.8158\n",
      "Epoch 17/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.5875 - accuracy: 0.7800 - val_loss: 0.7001 - val_accuracy: 0.7763\n",
      "Epoch 18/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6299 - accuracy: 0.7233 - val_loss: 0.7362 - val_accuracy: 0.6447\n",
      "Epoch 19/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.5956 - accuracy: 0.7700 - val_loss: 0.5274 - val_accuracy: 0.8026\n",
      "Epoch 20/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6067 - accuracy: 0.7600 - val_loss: 0.5623 - val_accuracy: 0.8026\n",
      "Epoch 21/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6247 - accuracy: 0.7367 - val_loss: 0.5745 - val_accuracy: 0.8158\n",
      "Epoch 22/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6710 - accuracy: 0.7467 - val_loss: 0.7488 - val_accuracy: 0.6842\n",
      "Epoch 23/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.7476 - accuracy: 0.7333 - val_loss: 0.7092 - val_accuracy: 0.7105\n",
      "Epoch 24/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6425 - accuracy: 0.7600 - val_loss: 0.6860 - val_accuracy: 0.6711\n",
      "Epoch 25/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.5599 - accuracy: 0.7867 - val_loss: 0.6215 - val_accuracy: 0.7763\n",
      "Epoch 26/60\n",
      "10/10 [==============================] - 0s 2ms/step - loss: 0.5601 - accuracy: 0.7467 - val_loss: 0.6121 - val_accuracy: 0.8026\n",
      "Epoch 27/60\n",
      "10/10 [==============================] - 0s 2ms/step - loss: 0.5304 - accuracy: 0.8067 - val_loss: 0.5019 - val_accuracy: 0.8421\n",
      "Epoch 28/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.5428 - accuracy: 0.7900 - val_loss: 0.6132 - val_accuracy: 0.7763\n",
      "Epoch 29/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6024 - accuracy: 0.7400 - val_loss: 0.5278 - val_accuracy: 0.8289\n",
      "Epoch 30/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4595 - accuracy: 0.8133 - val_loss: 0.5046 - val_accuracy: 0.7763\n",
      "Epoch 31/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4918 - accuracy: 0.8267 - val_loss: 0.4743 - val_accuracy: 0.7895\n",
      "Epoch 32/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4843 - accuracy: 0.7800 - val_loss: 0.4917 - val_accuracy: 0.8289\n",
      "Epoch 33/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4456 - accuracy: 0.8333 - val_loss: 0.5129 - val_accuracy: 0.7895\n",
      "Epoch 34/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3834 - accuracy: 0.8300 - val_loss: 0.5339 - val_accuracy: 0.8026\n",
      "Epoch 35/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4371 - accuracy: 0.8133 - val_loss: 0.3991 - val_accuracy: 0.8553\n",
      "Epoch 36/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4023 - accuracy: 0.8333 - val_loss: 0.5085 - val_accuracy: 0.8158\n",
      "Epoch 37/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3903 - accuracy: 0.8333 - val_loss: 0.4627 - val_accuracy: 0.8289\n",
      "Epoch 38/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3675 - accuracy: 0.8733 - val_loss: 0.4706 - val_accuracy: 0.8289\n",
      "Epoch 39/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3274 - accuracy: 0.8733 - val_loss: 0.5187 - val_accuracy: 0.8553\n",
      "Epoch 40/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 1.1791 - accuracy: 0.6600 - val_loss: 1.2795 - val_accuracy: 0.5263\n",
      "Epoch 41/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.9043 - accuracy: 0.5967 - val_loss: 0.7887 - val_accuracy: 0.7500\n",
      "Epoch 42/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.7107 - accuracy: 0.6767 - val_loss: 0.8335 - val_accuracy: 0.7368\n",
      "Epoch 43/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.6468 - accuracy: 0.7400 - val_loss: 0.6290 - val_accuracy: 0.8026\n",
      "Epoch 44/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.5384 - accuracy: 0.7733 - val_loss: 0.6434 - val_accuracy: 0.7237\n",
      "Epoch 45/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4648 - accuracy: 0.8133 - val_loss: 0.5061 - val_accuracy: 0.8421\n",
      "Epoch 46/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3998 - accuracy: 0.8767 - val_loss: 0.4923 - val_accuracy: 0.8553\n",
      "Epoch 47/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3556 - accuracy: 0.8800 - val_loss: 0.4067 - val_accuracy: 0.8684\n",
      "Epoch 48/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4314 - accuracy: 0.8167 - val_loss: 0.6064 - val_accuracy: 0.7237\n",
      "Epoch 49/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3928 - accuracy: 0.8367 - val_loss: 0.4709 - val_accuracy: 0.8026\n",
      "Epoch 50/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4070 - accuracy: 0.8533 - val_loss: 0.6085 - val_accuracy: 0.7368\n",
      "Epoch 51/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.4946 - accuracy: 0.8333 - val_loss: 0.3807 - val_accuracy: 0.8553\n",
      "Epoch 52/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3722 - accuracy: 0.8533 - val_loss: 0.5308 - val_accuracy: 0.7632\n",
      "Epoch 53/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3505 - accuracy: 0.8733 - val_loss: 0.5116 - val_accuracy: 0.7763\n",
      "Epoch 54/60\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3515 - accuracy: 0.8533 - val_loss: 0.4665 - val_accuracy: 0.8421\n",
      "Epoch 55/60\n",
      "10/10 [==============================] - 0s 2ms/step - loss: 0.3743 - accuracy: 0.8433 - val_loss: 0.3766 - val_accuracy: 0.8684\n",
      "Epoch 56/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3547 - accuracy: 0.8700 - val_loss: 0.5346 - val_accuracy: 0.7763\n",
      "Epoch 57/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3670 - accuracy: 0.8400 - val_loss: 0.4724 - val_accuracy: 0.7632\n",
      "Epoch 58/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.3415 - accuracy: 0.8367 - val_loss: 0.3843 - val_accuracy: 0.8289\n",
      "Epoch 59/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.2773 - accuracy: 0.8800 - val_loss: 0.4036 - val_accuracy: 0.8421\n",
      "Epoch 60/60\n",
      "10/10 [==============================] - 0s 3ms/step - loss: 0.2707 - accuracy: 0.9133 - val_loss: 0.4254 - val_accuracy: 0.8421\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x1b2d3b60790>"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X_train, y_train, verbose=1, validation_data=(X_val, y_val), epochs=60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "PghHskAJXSzf",
    "outputId": "5808ce33-c956-448e-90a1-49ccbe24db5f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[24,  0,  1,  0,  0],\n",
       "       [ 0, 10,  4,  1,  0],\n",
       "       [ 0,  0, 15,  0,  0],\n",
       "       [ 0,  0,  0, 11,  0],\n",
       "       [ 0,  8,  2,  0, 18]], dtype=int64)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = model.predict(X_test)\n",
    "y_pred_classes = np.argmax(y_pred, axis=1)\n",
    "confusion_matrix(y_test, y_pred_classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/"
    },
    "id": "E6XEsmp0XYzK",
    "outputId": "bce8ded7-d215-4b16-a1fe-a7fc7cf9fb2a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      0.96      0.98        25\n",
      "           1       0.56      0.67      0.61        15\n",
      "           2       0.68      1.00      0.81        15\n",
      "           3       0.92      1.00      0.96        11\n",
      "           4       1.00      0.64      0.78        28\n",
      "\n",
      "    accuracy                           0.83        94\n",
      "   macro avg       0.83      0.85      0.83        94\n",
      "weighted avg       0.87      0.83      0.83        94\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_test, y_pred_classes))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "id": "dnCdoBHlghPY"
   },
   "outputs": [],
   "source": [
    "#Saving the model\n",
    "model_json = model.to_json()\n",
    "with open(\"model.json\", \"w\") as json_file:\n",
    "    json_file.write(model_json)\n",
    "# serialize weights to HDF5\n",
    "model.save_weights(\"model.h5\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "provenance": []
  },
  "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": "3.8.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}