diff --git a/AdultDataset.ipynb b/AdultDataset.ipynb
index a84e634..175a92c 100644
--- a/AdultDataset.ipynb
+++ b/AdultDataset.ipynb
@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 120,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -11,7 +11,10 @@
     "import numpy as np\n",
     "import torch\n",
     "from sklearn.model_selection import KFold, StratifiedKFold\n",
-    "import math"
+    "import math\n",
+    "import seaborn as sns\n",
+    "import matplotlib.pyplot as plt\n",
+    "plt.rcParams[\"figure.figsize\"] = (16, 9)"
    ]
   },
   {
@@ -36,7 +39,7 @@
     "    DecisionTreeClassifier(),\n",
     "    RandomForestClassifier(n_jobs=4),\n",
     "    AdaBoostClassifier()\n",
-    "]"
+    "]\n"
    ]
   },
   {
@@ -51,7 +54,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 29,
+   "execution_count": 151,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -76,7 +79,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 152,
    "metadata": {},
    "outputs": [
     {
@@ -85,7 +88,7 @@
        "(45222, 15)"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 152,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -98,7 +101,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 153,
    "metadata": {},
    "outputs": [
     {
@@ -107,7 +110,7 @@
        "(array(['<=50K', '>50K'], dtype=object), array([34014, 11208]))"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 153,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -118,7 +121,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 154,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -127,7 +130,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 155,
    "metadata": {},
    "outputs": [
     {
@@ -166,7 +169,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 156,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -184,7 +187,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 157,
    "metadata": {},
    "outputs": [
     {
@@ -211,7 +214,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 158,
    "metadata": {},
    "outputs": [
     {
@@ -244,7 +247,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 159,
    "metadata": {},
    "outputs": [
     {
@@ -270,7 +273,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": 160,
    "metadata": {},
    "outputs": [
     {
@@ -338,7 +341,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 161,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -350,7 +353,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 162,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -361,7 +364,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 163,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -370,7 +373,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 164,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -379,7 +382,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 165,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -398,7 +401,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 166,
    "metadata": {},
    "outputs": [
     {
@@ -417,7 +420,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 45,
+   "execution_count": 167,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -438,7 +441,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 53,
+   "execution_count": 168,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -454,7 +457,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 74,
+   "execution_count": 169,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -481,90 +484,178 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
+   "execution_count": 185,
+   "metadata": {
+    "scrolled": false
+   },
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "Noise amount:   0%|          | 0/11 [00:00<?, ?it/s]\n",
+      "Noise amount:   0%|          | 0/8 [00:00<?, ?it/s]\n",
+      "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:01<00:02,  1.28s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [00:09<00:03,  3.29s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [00:21<00:00,  7.04s/it]\u001b[A\n",
+      "\n",
+      "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:01<00:02,  1.34s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [00:12<00:04,  4.18s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [00:24<00:00,  8.23s/it]\u001b[A\n",
+      "Noise amount:  12%|█▎        | 1/8 [00:45<05:20, 45.80s/it]\n",
+      "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:40<01:20, 40.34s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [03:15<01:14, 74.64s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [05:02<00:00, 100.72s/it][A\n",
+      "\n",
+      "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:51<01:43, 51.57s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [04:04<01:33, 94.00s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [05:52<00:00, 117.52s/it][A\n",
+      "Noise amount:  25%|██▌       | 2/8 [12:10<23:45, 237.52s/it]\n",
+      "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:28<00:56, 28.50s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:19<00:53, 53.40s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:11<00:00, 83.79s/it]\u001b[A\n",
+      "\n",
+      "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:47<01:35, 47.77s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [03:43<01:26, 86.21s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [05:20<00:00, 106.97s/it][A\n",
+      "Noise amount:  38%|███▊      | 3/8 [22:14<28:57, 347.56s/it]\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:01<00:02,  1.23s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [00:08<00:02,  3.00s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [00:17<00:00,  5.93s/it]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:25<00:51, 25.57s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:01<00:46, 46.78s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [03:49<00:00, 76.60s/it]\u001b[A\n",
       "\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:01<00:02,  1.26s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [00:11<00:03,  3.94s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [00:21<00:00,  7.20s/it]\u001b[A\n",
-      "Noise amount:   9%|▉         | 1/11 [00:39<06:33, 39.39s/it]\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:38<01:16, 38.27s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [03:00<01:09, 69.54s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:58<00:00, 99.60s/it]\u001b[A\n",
+      "Noise amount:  50%|█████     | 4/8 [31:33<27:23, 410.88s/it]\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:39<01:19, 39.65s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [03:05<01:11, 71.58s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [04:38<00:00, 92.73s/it]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:26<00:53, 26.91s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:18<00:52, 52.17s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:09<00:00, 83.33s/it]\u001b[A\n",
       "\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:52<01:44, 52.22s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [03:53<01:30, 90.99s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [05:29<00:00, 109.88s/it][A\n",
-      "Noise amount:  18%|█▊        | 2/11 [11:14<32:43, 218.12s/it]\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:36<01:13, 36.83s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:52<01:06, 66.51s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:56<00:00, 98.76s/it]\u001b[A\n",
+      "Noise amount:  62%|██████▎   | 5/8 [41:08<23:00, 460.03s/it]\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:25<00:51, 25.98s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [02:09<00:49, 49.34s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [03:42<00:00, 74.21s/it]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:38<01:16, 38.08s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:27<00:59, 59.52s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:30<00:00, 90.10s/it]\u001b[A\n",
       "\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:38<01:17, 38.85s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [02:58<01:09, 69.15s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [04:33<00:00, 91.04s/it]\u001b[A\n",
-      "Noise amount:  27%|██▋       | 3/11 [19:58<41:18, 309.78s/it]\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:37<01:14, 37.03s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:47<01:04, 64.94s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:53<00:00, 97.86s/it]\u001b[A\n",
+      "Noise amount:  75%|███████▌  | 6/8 [51:08<16:44, 502.10s/it]\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:24<00:48, 24.41s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [01:57<00:45, 45.11s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [03:30<00:00, 70.14s/it]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:37<01:15, 37.99s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:37<01:02, 62.49s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:45<00:00, 95.26s/it]\u001b[A\n",
       "\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:33<01:06, 33.00s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [03:30<01:16, 76.47s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [05:25<00:00, 108.62s/it][A\n",
-      "Noise amount:  36%|███▋      | 4/11 [29:21<45:01, 385.92s/it]\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:41<01:23, 41.63s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:58<01:10, 70.32s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:50<00:00, 96.78s/it]\u001b[A\n",
+      "Noise amount:  88%|████████▊ | 7/8 [1:01:25<08:56, 536.69s/it]\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:30<01:01, 30.78s/it]\u001b[A\n",
-      "\tClassifier:  67%|██████▋   | 2/3 [02:11<00:51, 51.66s/it]\u001b[A\n",
-      "\tClassifier: 100%|██████████| 3/3 [03:58<00:00, 79.49s/it]\u001b[A\n",
+      "\tClassifier:  33%|███▎      | 1/3 [00:32<01:05, 32.67s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:26<00:57, 57.11s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:08<00:00, 82.93s/it]\u001b[A\n",
       "\n",
       "\tClassifier:   0%|          | 0/3 [00:00<?, ?it/s]\u001b[A\n",
-      "\tClassifier:  33%|███▎      | 1/3 [00:33<01:06, 33.30s/it]\u001b[A"
+      "\tClassifier:  33%|███▎      | 1/3 [00:38<01:16, 38.31s/it]\u001b[A\n",
+      "\tClassifier:  67%|██████▋   | 2/3 [02:51<01:06, 66.90s/it]\u001b[A\n",
+      "\tClassifier: 100%|██████████| 3/3 [04:41<00:00, 93.92s/it]\u001b[A\n",
+      "Noise amount: 100%|██████████| 8/8 [1:10:45<00:00, 530.72s/it]\n"
      ]
     }
    ],
    "source": [
-    "scores = []\n",
-    "for eps in tqdm.tqdm(['NA'] + np.arange(0.1, 5, 0.5).tolist(), desc='Noise amount'):\n",
+    "scores = [['Epsilon, Utility_Privacy', 'Classifier', 'Precision', 'Recall', 'F1', 'Accuracy']]\n",
+    "for eps in tqdm.tqdm(['NA', 0.1, 0.5, 1, 2, 4, 7, 10], desc='Noise amount'):\n",
     "    for attr in ['class', 'workclass']:\n",
-    "        num_cls = 2 if attr=='class' else 7\n",
+    "        num_cls = 2 if attr=='class' else 4\n",
     "        scores += run(df_x, num_cls, attr, eps=eps)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 191,
    "metadata": {},
    "outputs": [],
    "source": [
-    "import pickle\n",
-    "\n",
-    "with open('noisy_scores.dat', 'wb') as fp:\n",
-    "    pickle.dump(itemlist, fp)"
+    "scores[0] = ['Epsilon', 'Utility_Privacy',\n",
+    "  'Classifier',\n",
+    "  'Precision',\n",
+    "  'Recall',\n",
+    "  'F1', 'Accuracy']"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 192,
    "metadata": {},
    "outputs": [],
-   "source": []
+   "source": [
+    "sdf = pd.DataFrame(columns=scores[0], data=scores[1:])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 193,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1152x648 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.boxplot(x='Utility_Privacy', y='Accuracy', data=sdf, hue='Epsilon')\n",
+    "plt.title('Accuracy')\n",
+    "plt.savefig('Accuracy.png',bbox_inches = 'tight',\n",
+    "    pad_inches = 0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 194,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1152x648 with 1 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "sns.boxplot(x='Utility_Privacy', y='Recall', data=sdf, hue='Epsilon')\n",
+    "plt.title('Recall')\n",
+    "plt.savefig('Sensitivity.png', bbox_inches = 'tight',\n",
+    "    pad_inches = 0)"
+   ]
   },
   {
    "cell_type": "code",