Machine Learning

Please see Question.pdf

• Question.pdf

This material may consist of step-by-step explanations on how to solve a problem or examples of proper writing, including the use of citations, references, bibliographies, and formatting. This material is made available for the sole purpose of studying and learning - misuse is strictly forbidden.

],
   "source": [
    "X, y = make_blobs(n_samples=300, centers =3, cluster_std=3.5,random_state=42)\n",
    "plt.figure(figsize=(8, 6))\n",
    "ax=plt.subplot(111)\n",
    "ax.scatter(X[:, 0], X[:, 1], c=y, s=40, cmap='viridis')\n",
    "plt.show()"
   ]
},
{
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "URtzpJITZ04e"
   },
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.4,\n",
    "random_state=0)"
   ]
},
{
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "MHhIZTnVba0U"
   },
   "source": [
    "## 1."
   ]
},
{
   "cell_type": "code",
   "execution_count": 0,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "MNNnB5IAbaUA"
   },
   "outputs": [],
   "source": [
    "k = 3\n",
    "\n",
    "epsilon = 0.00001\n",
    "max_iter = 10000\n",
    "\n",
    "\n",
    "def dist(row_a, row_b):\n",
    "    d = 0\n",
    "    for i in range(len(row_a)):\n",
    "       d += (row_a[i] - row_b[i]) ** 2\n",
    "    d = math.sqrt(d)\n",
    "    return d\n",
    "\n",
    "df = pd.DataFrame(X_train)\n",
    "df.reset_index()\n",
    "\n",
    "n_test = len(X_test)\n",
    "\n",
    "df[\"pseudo\"] = y_train\n",
    "\n",
    "centroids = df.groupby(\"pseudo\").agg(np.mean)\n",
    "centroids.reset_index()\n",
    "\n",
    "\n",
    "index = 0\n",
    "df_test = None\n",
    "while index < max_iter:\n",
    "    pseudo_labels = []\n",
    "    for i in range(0, n_test):\n",
    "       dis = []\n",
    "       for cluster in range(0, k):\n",
    "            d = dist(centroids.iloc[cluster,], X_test[i])\n",
    "            dis.append(d)\n",
    "       pseudo_label = dis.index(min(dis))\n",
    "       pseudo_labels.append(pseudo_label)\n",
    "\n",
    "    df_test = pd.DataFrame(X_test)\n",
    "    df_test[\"pseudo\"] = pseudo_labels\n",
    "    df3 = df.append(df_test, ignore_index=True)\n",
    "    centroids_new = df3.groupby(\"pseudo\").agg(np.mean)\n",
    "    centroids_new.reset_index()\n",
    "    all_converged = 0\n",
    "    for i in range(0, len(centroids)):\n",
    "       d = dist(centroids.iloc[i, ], centroids_new.iloc[i,])\n",
    "       if d < epsilon:\n",
    "            all_converged += 1\n",
    "    if all_converged == k:\n",
    "       break\n",
    "    else:\n",
    "       centroids = centroids_new\n",
    "\n",
    "    index += 1\n",
    "df_test[\"y\"] = y_test\n",
    "df[\"y\"] = y_train\n",
    "\n"
   ]
},
{
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "rFYcDlko-5Tq"
   },
   "source": [
    "## 2."
   ]
},
{
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {
    "colab": {
    "base_uri": "https://localhost:8080/",
    "height": 190
    },
    "colab_type": "code",
    "id": "NN6hYscT8ctk",
    "outputId": "def713ac-5095-4e15-de09-9e12f2b663be"
   },
   "outputs": [
    {
    "name": "stdout",
    "output_type": "stream",
    "text": [
      "Mean of each cluster:\n"
    ]
    },
    {
    "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>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>y</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       " </thead>\n",
       " <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-2.836760</td>\n",
       "      <td>9.192985</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.654930</td>\n",
       "      <td>2.271495</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-6.839723</td>\n",
       "      <td>-7.237148</td>\n",
       "    </tr>\n",
       " </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          0         1\n",
       "y                   \n",
       "0 -2.836760 9.192985\n",
       "1 4.654930 2.271495\n",
       "2 -6.839723 -7.237148"
      ]
    },
    "execution_count": 42,
    "metadata": {
      "tags": []
    },
    "output_type": "execute_result"
    }
   ],
   "source": [
    "test_pseudo = df_test[\"pseudo\"]\n",
    "del df_test[\"pseudo\"]\n",
    "df_test.reset_index()\n",
    "mean_mat = df_test.groupby(\"y\").agg(np.mean)\n",
    "print(\"Mean of each cluster:\")\n",
    "mean_mat"
   ]
},
{
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {
    "base_uri": "https://localhost:8080/",
    "height": 190
    },
    "colab_type": "code",
    "id": "aQvDj9dT8p3r",
    "outputId": "22cd8019-8c58-4cf2-ee74-f74ad790eb86"
   },
   "outputs": [
    {
    "name": "stdout...