diff --git a/Train.ipynb b/Train.ipynb
index f3ed067..f582391 100644
--- a/Train.ipynb
+++ b/Train.ipynb
@@ -15,7 +15,7 @@
     "\n",
     "import config\n",
     "import utils\n",
-    "import main_classificaiton as mc"
+    "import main_regression as mr"
    ]
   },
   {
@@ -37,28 +37,65 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 21,
+   "id": "877996be",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import glob\n",
+    "import os\n",
+    "\n",
+    "def Saver(model_name):\n",
+    "\n",
+    "    os.makedirs('./ckpt', exist_ok=True)\n",
+    "    os.makedirs('./ckpt/' + model_name, exist_ok=True)\n",
+    "    \n",
+    "    model_dir_path = './ckpt/' + model_name\n",
+    "    runs = sorted(glob.glob(os.path.join('./', model_name, '/experiment_*')))\n",
+    "    indices = []\n",
+    "    for tmp in runs:\n",
+    "        tmp_num = tmp.split(\"\\\\\")[-1]\n",
+    "        tmp_num = int(tmp_num.split(\"_\")[-1])\n",
+    "        indices.append(tmp_num)\n",
+    "        \n",
+    "    if len(indices) == 0:\n",
+    "        run_id = str(0)\n",
+    "    else:\n",
+    "        run_id = np.max(indices) + 1\n",
+    "\n",
+    "    experiment_dir = os.path.join(model_dir_path, '/', f'experiment_{str(run_id)}')\n",
+    "\n",
+    "    os.makedirs(experiment_dir, exist_ok = True)\n",
+    "    \n",
+    "\n",
+    "    return experiment_dir"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "(7352, 9, 128)\n",
-      "(7352,)\n",
-      "(2947, 9, 128)\n",
-      "(2947,)\n",
-      "inputSize(train_x.shape[1]): 9\n",
-      "sequenceLenth (train_x.shape[2]): 128\n",
-      "Save MinMaxScaler in path: ./scaler/minmax_scaler_x.pkl\n"
+      "(95, 24, 144)\n",
+      "(95,)\n",
+      "(42, 24, 144)\n",
+      "(42,)\n",
+      "inputSize(train_x.shape[1]): 24\n",
+      "sequenceLenth (train_x.shape[2]): 144\n",
+      "Save MinMaxScaler in path: ./scaler/minmax_scaler_x.pkl\n",
+      "Save MinMaxScaler in path: ./scaler/minmax_scaler_y.pkl\n"
      ]
     }
    ],
    "source": [
     "# load raw data\n",
-    "# [\"LSTM\", \"GRU\", \"CNN_1D\", \"LSTM_FCNs\"]는 사용 데이터가 동일하기 때문에 편의상 utils.load_data에서 model_name을 'LSTM'로 설정하여 불러온 데이터를 함께 사용함\n",
+    "# [\"LSTM_rg\", \"GRU_rg\", \"CNN_1D_rg\", \"LSTM_FCNs_rg\"]는 사용 데이터가 동일하기 때문에 편의상 utils.load_data에서 model_name을 'LSTM_rg'로 설정하여 불러온 데이터를 함께 사용함\n",
     "data_root_dir = './data/'\n",
-    "train_x, train_y, test_x, test_y = utils.load_data(data_root_dir, model_name='LSTM')  # shape=(num_of_instance, input_dims, time_steps)\n",
+    "train_x, train_y, test_x, test_y = utils.load_data(data_root_dir, model_name='LSTM_rg')  # shape=(num_of_instance, input_dims, time_steps)\n",
     "\n",
     "# split train data into train/valiation data\n",
     "# train data를 랜덤으로 test_size=split_ratio에 대하여 train/validation set으로 분할 (관측치 단위 데이터)\n",
@@ -67,12 +104,14 @@
     "\n",
     "# normalization\n",
     "scaler_x_path = './scaler/minmax_scaler_x.pkl'\n",
-    "train_x, valid_x = utils.get_train_val_data(train_x, valid_x, scaler_x_path)"
+    "scaler_y_path = './scaler/minmax_scaler_y.pkl'\n",
+    "train_x, valid_x = utils.get_train_val_data(train_x, valid_x, scaler_x_path)\n",
+    "train_y, valid_y = utils.get_train_val_data(train_y, valid_y, scaler_y_path)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 24,
    "metadata": {
     "scrolled": true
    },
@@ -81,85 +120,34 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Start training model: LSTM\n",
-      "\n",
-      "Epoch 1/150\n",
-      "train Loss: 1.7860 Acc: 0.2268\n",
-      "val Loss: 1.7782 Acc: 0.1903\n",
-      "\n",
-      "Epoch 10/150\n",
-      "train Loss: 1.0381 Acc: 0.5348\n",
-      "val Loss: 1.0081 Acc: 0.5364\n",
-      "\n",
-      "Epoch 20/150\n",
-      "train Loss: 0.9961 Acc: 0.5482\n",
-      "val Loss: 1.0043 Acc: 0.5398\n",
-      "\n",
-      "Epoch 30/150\n",
-      "train Loss: 0.9771 Acc: 0.5533\n",
-      "val Loss: 0.9687 Acc: 0.5574\n",
-      "\n",
-      "Epoch 40/150\n",
-      "train Loss: 0.9759 Acc: 0.5599\n",
-      "val Loss: 0.9543 Acc: 0.5554\n",
-      "\n",
-      "Epoch 50/150\n",
-      "train Loss: 0.9517 Acc: 0.5666\n",
-      "val Loss: 0.9388 Acc: 0.5697\n",
-      "\n",
-      "Epoch 60/150\n",
-      "train Loss: 0.9573 Acc: 0.5698\n",
-      "val Loss: 0.9812 Acc: 0.5534\n",
-      "\n",
-      "Epoch 70/150\n",
-      "train Loss: 0.9433 Acc: 0.5718\n",
-      "val Loss: 0.9271 Acc: 0.5717\n",
-      "\n",
-      "Epoch 80/150\n",
-      "train Loss: 0.9461 Acc: 0.5739\n",
-      "val Loss: 0.9663 Acc: 0.5506\n",
-      "\n",
-      "Epoch 90/150\n",
-      "train Loss: 0.9134 Acc: 0.5980\n",
-      "val Loss: 0.8862 Acc: 0.5908\n",
-      "\n",
-      "Epoch 100/150\n",
-      "train Loss: 0.8907 Acc: 0.6058\n",
-      "val Loss: 0.8716 Acc: 0.6071\n",
+      "Start training model: LSTM_rg\n",
       "\n",
-      "Epoch 110/150\n",
-      "train Loss: 0.9150 Acc: 0.5990\n",
-      "val Loss: 0.9410 Acc: 0.5778\n",
+      "Epoch 1/100\n",
+      "train Loss: 0.0566\n",
+      "val Loss: 0.0569\n",
       "\n",
-      "Epoch 120/150\n",
-      "train Loss: 0.4736 Acc: 0.8242\n",
-      "val Loss: 0.4536 Acc: 0.8307\n",
+      "Epoch 50/100\n",
+      "train Loss: 0.0434\n",
+      "val Loss: 0.0434\n",
       "\n",
-      "Epoch 130/150\n",
-      "train Loss: 0.3744 Acc: 0.8536\n",
-      "val Loss: 0.3544 Acc: 0.8511\n",
+      "Epoch 100/100\n",
+      "train Loss: 0.0326\n",
+      "val Loss: 0.0346\n",
       "\n",
-      "Epoch 140/150\n",
-      "train Loss: 0.3119 Acc: 0.8786\n",
-      "val Loss: 0.3006 Acc: 0.8749\n",
-      "\n",
-      "Epoch 150/150\n",
-      "train Loss: 0.2581 Acc: 0.8978\n",
-      "val Loss: 0.2512 Acc: 0.8973\n",
-      "\n",
-      "Training complete in 2m 8s\n",
-      "Best val Acc: 0.899388\n"
+      "Training complete in 0m 5s\n",
+      "Best val MSE: 0.034624\n"
      ]
     }
    ],
    "source": [
     "# Case 1. LSTM model (w/o data representation)\n",
-    "model_name = 'LSTM'\n",
+    "model_name = 'LSTM_rg'\n",
     "model_params = config.model_config[model_name]\n",
     "\n",
-    "data_cls = mc.Classification(model_params)\n",
-    "best_model = data_cls.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
-    "data_cls.save_model(best_model, best_model_path=model_params[\"best_model_path\"])  # 모델 저장"
+    "exp_path = Saver(model_name)\n",
+    "data_reg = mr.Regression(model_params)\n",
+    "best_model = data_reg.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
+    "data_reg.save_model(best_model, best_model_path=os.path.join(exp_path, model_params[\"best_model\"]))  # 모델 저장"
    ]
   },
   {
@@ -173,85 +161,106 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Start training model: GRU\n",
+      "Start training model: GRU_rg\n",
+      "\n",
+      "Epoch 1/1000\n",
+      "train Loss: 0.0489\n",
+      "val Loss: 0.0434\n",
+      "\n",
+      "Epoch 50/1000\n",
+      "train Loss: 0.0342\n",
+      "val Loss: 0.0434\n",
+      "\n",
+      "Epoch 100/1000\n",
+      "train Loss: 0.0218\n",
+      "val Loss: 0.0316\n",
+      "\n",
+      "Epoch 150/1000\n",
+      "train Loss: 0.0187\n",
+      "val Loss: 0.0232\n",
       "\n",
-      "Epoch 1/150\n",
-      "train Loss: 1.7731 Acc: 0.1916\n",
-      "val Loss: 1.7537 Acc: 0.1903\n",
+      "Epoch 200/1000\n",
+      "train Loss: 0.0160\n",
+      "val Loss: 0.0187\n",
       "\n",
-      "Epoch 10/150\n",
-      "train Loss: 1.0544 Acc: 0.5222\n",
-      "val Loss: 1.0374 Acc: 0.5255\n",
+      "Epoch 250/1000\n",
+      "train Loss: 0.0149\n",
+      "val Loss: 0.0210\n",
       "\n",
-      "Epoch 20/150\n",
-      "train Loss: 1.0031 Acc: 0.5485\n",
-      "val Loss: 0.9924 Acc: 0.5472\n",
+      "Epoch 300/1000\n",
+      "train Loss: 0.0141\n",
+      "val Loss: 0.0207\n",
       "\n",
-      "Epoch 30/150\n",
-      "train Loss: 0.9822 Acc: 0.5562\n",
-      "val Loss: 0.9828 Acc: 0.5520\n",
+      "Epoch 350/1000\n",
+      "train Loss: 0.0120\n",
+      "val Loss: 0.0214\n",
       "\n",
-      "Epoch 40/150\n",
-      "train Loss: 0.9771 Acc: 0.5650\n",
-      "val Loss: 0.9612 Acc: 0.5602\n",
+      "Epoch 400/1000\n",
+      "train Loss: 0.0104\n",
+      "val Loss: 0.0217\n",
       "\n",
-      "Epoch 50/150\n",
-      "train Loss: 0.9622 Acc: 0.5676\n",
-      "val Loss: 0.9475 Acc: 0.5765\n",
+      "Epoch 450/1000\n",
+      "train Loss: 0.0092\n",
+      "val Loss: 0.0194\n",
       "\n",
-      "Epoch 60/150\n",
-      "train Loss: 0.9493 Acc: 0.5732\n",
-      "val Loss: 0.9281 Acc: 0.5819\n",
+      "Epoch 500/1000\n",
+      "train Loss: 0.0081\n",
+      "val Loss: 0.0173\n",
       "\n",
-      "Epoch 70/150\n",
-      "train Loss: 0.9316 Acc: 0.5834\n",
-      "val Loss: 0.9118 Acc: 0.5887\n",
+      "Epoch 550/1000\n",
+      "train Loss: 0.0074\n",
+      "val Loss: 0.0167\n",
       "\n",
-      "Epoch 80/150\n",
-      "train Loss: 0.9022 Acc: 0.6007\n",
-      "val Loss: 0.8824 Acc: 0.6044\n",
+      "Epoch 600/1000\n",
+      "train Loss: 0.0066\n",
+      "val Loss: 0.0169\n",
       "\n",
-      "Epoch 90/150\n",
-      "train Loss: 0.6534 Acc: 0.7789\n",
-      "val Loss: 0.6194 Acc: 0.7587\n",
+      "Epoch 650/1000\n",
+      "train Loss: 0.0058\n",
+      "val Loss: 0.0166\n",
       "\n",
-      "Epoch 100/150\n",
-      "train Loss: 0.3516 Acc: 0.8764\n",
-      "val Loss: 0.3273 Acc: 0.8797\n",
+      "Epoch 700/1000\n",
+      "train Loss: 0.0074\n",
+      "val Loss: 0.0194\n",
       "\n",
-      "Epoch 110/150\n",
-      "train Loss: 0.2687 Acc: 0.9004\n",
-      "val Loss: 0.2482 Acc: 0.9062\n",
+      "Epoch 750/1000\n",
+      "train Loss: 0.0053\n",
+      "val Loss: 0.0207\n",
       "\n",
-      "Epoch 120/150\n",
-      "train Loss: 0.2203 Acc: 0.9162\n",
-      "val Loss: 0.2025 Acc: 0.9198\n",
+      "Epoch 800/1000\n",
+      "train Loss: 0.0043\n",
+      "val Loss: 0.0209\n",
       "\n",
-      "Epoch 130/150\n",
-      "train Loss: 0.1857 Acc: 0.9259\n",
-      "val Loss: 0.1803 Acc: 0.9252\n",
+      "Epoch 850/1000\n",
+      "train Loss: 0.0041\n",
+      "val Loss: 0.0283\n",
       "\n",
-      "Epoch 140/150\n",
-      "train Loss: 0.1727 Acc: 0.9306\n",
-      "val Loss: 0.1688 Acc: 0.9279\n",
+      "Epoch 900/1000\n",
+      "train Loss: 0.0038\n",
+      "val Loss: 0.0195\n",
       "\n",
-      "Epoch 150/150\n",
-      "train Loss: 0.1578 Acc: 0.9359\n",
-      "val Loss: 0.1551 Acc: 0.9361\n",
+      "Epoch 950/1000\n",
+      "train Loss: 0.0052\n",
+      "val Loss: 0.0222\n",
       "\n",
-      "Training complete in 2m 5s\n",
-      "Best val Acc: 0.936098\n"
+      "Epoch 1000/1000\n",
+      "train Loss: 0.0034\n",
+      "val Loss: 0.0215\n",
+      "\n",
+      "Training complete in 0m 50s\n",
+      "Best val MSE: 0.013994\n"
      ]
     }
    ],
    "source": [
     "# Case 2. GRU (w/o data representation)\n",
-    "model_name = 'GRU'\n",
+    "model_name = 'GRU_rg'\n",
     "model_params = config.model_config[model_name]\n",
     "\n",
-    "data_cls = mc.Classification(model_params)\n",
-    "best_model = data_cls.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
-    "data_cls.save_model(best_model, best_model_path=model_params[\"best_model_path\"])  # 모델 저장"
+    "exp_path = Saver(model_name)\n",
+    "data_reg = mr.Regression(model_params)\n",
+    "best_model = data_reg.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
+    "data_reg.save_model(best_model, best_model_path=os.path.join(exp_path, model_params[\"best_model\"]))  # 모델 저장"
    ]
   },
   {
@@ -265,85 +274,106 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Start training model: CNN_1D\n",
+      "Start training model: CNN_1D_rg\n",
+      "\n",
+      "Epoch 1/1000\n",
+      "train Loss: 0.0499\n",
+      "val Loss: 0.0436\n",
+      "\n",
+      "Epoch 50/1000\n",
+      "train Loss: 0.0363\n",
+      "val Loss: 0.0441\n",
+      "\n",
+      "Epoch 100/1000\n",
+      "train Loss: 0.0315\n",
+      "val Loss: 0.0448\n",
+      "\n",
+      "Epoch 150/1000\n",
+      "train Loss: 0.0270\n",
+      "val Loss: 0.0450\n",
       "\n",
-      "Epoch 1/150\n",
-      "train Loss: 1.7472 Acc: 0.3323\n",
-      "val Loss: 1.6865 Acc: 0.3494\n",
+      "Epoch 200/1000\n",
+      "train Loss: 0.0238\n",
+      "val Loss: 0.0458\n",
       "\n",
-      "Epoch 10/150\n",
-      "train Loss: 0.8122 Acc: 0.7369\n",
-      "val Loss: 0.7855 Acc: 0.7362\n",
+      "Epoch 250/1000\n",
+      "train Loss: 0.0187\n",
+      "val Loss: 0.0479\n",
       "\n",
-      "Epoch 20/150\n",
-      "train Loss: 0.5346 Acc: 0.8164\n",
-      "val Loss: 0.5212 Acc: 0.8015\n",
+      "Epoch 300/1000\n",
+      "train Loss: 0.0160\n",
+      "val Loss: 0.0483\n",
       "\n",
-      "Epoch 30/150\n",
-      "train Loss: 0.4605 Acc: 0.8235\n",
-      "val Loss: 0.4499 Acc: 0.8253\n",
+      "Epoch 350/1000\n",
+      "train Loss: 0.0132\n",
+      "val Loss: 0.0493\n",
       "\n",
-      "Epoch 40/150\n",
-      "train Loss: 0.4274 Acc: 0.8322\n",
-      "val Loss: 0.4142 Acc: 0.8368\n",
+      "Epoch 400/1000\n",
+      "train Loss: 0.0107\n",
+      "val Loss: 0.0495\n",
       "\n",
-      "Epoch 50/150\n",
-      "train Loss: 0.4071 Acc: 0.8391\n",
-      "val Loss: 0.3944 Acc: 0.8498\n",
+      "Epoch 450/1000\n",
+      "train Loss: 0.0083\n",
+      "val Loss: 0.0484\n",
       "\n",
-      "Epoch 60/150\n",
-      "train Loss: 0.3847 Acc: 0.8444\n",
-      "val Loss: 0.3875 Acc: 0.8287\n",
+      "Epoch 500/1000\n",
+      "train Loss: 0.0071\n",
+      "val Loss: 0.0490\n",
       "\n",
-      "Epoch 70/150\n",
-      "train Loss: 0.3733 Acc: 0.8507\n",
-      "val Loss: 0.3694 Acc: 0.8477\n",
+      "Epoch 550/1000\n",
+      "train Loss: 0.0061\n",
+      "val Loss: 0.0485\n",
       "\n",
-      "Epoch 80/150\n",
-      "train Loss: 0.3632 Acc: 0.8570\n",
-      "val Loss: 0.3612 Acc: 0.8484\n",
+      "Epoch 600/1000\n",
+      "train Loss: 0.0049\n",
+      "val Loss: 0.0483\n",
       "\n",
-      "Epoch 90/150\n",
-      "train Loss: 0.3534 Acc: 0.8575\n",
-      "val Loss: 0.3512 Acc: 0.8518\n",
+      "Epoch 650/1000\n",
+      "train Loss: 0.0038\n",
+      "val Loss: 0.0495\n",
       "\n",
-      "Epoch 100/150\n",
-      "train Loss: 0.3443 Acc: 0.8641\n",
-      "val Loss: 0.3343 Acc: 0.8654\n",
+      "Epoch 700/1000\n",
+      "train Loss: 0.0034\n",
+      "val Loss: 0.0503\n",
       "\n",
-      "Epoch 110/150\n",
-      "train Loss: 0.3330 Acc: 0.8706\n",
-      "val Loss: 0.3309 Acc: 0.8640\n",
+      "Epoch 750/1000\n",
+      "train Loss: 0.0034\n",
+      "val Loss: 0.0501\n",
       "\n",
-      "Epoch 120/150\n",
-      "train Loss: 0.3227 Acc: 0.8738\n",
-      "val Loss: 0.3161 Acc: 0.8708\n",
+      "Epoch 800/1000\n",
+      "train Loss: 0.0028\n",
+      "val Loss: 0.0518\n",
       "\n",
-      "Epoch 130/150\n",
-      "train Loss: 0.3161 Acc: 0.8743\n",
-      "val Loss: 0.3134 Acc: 0.8613\n",
+      "Epoch 850/1000\n",
+      "train Loss: 0.0021\n",
+      "val Loss: 0.0517\n",
       "\n",
-      "Epoch 140/150\n",
-      "train Loss: 0.3044 Acc: 0.8806\n",
-      "val Loss: 0.3083 Acc: 0.8640\n",
+      "Epoch 900/1000\n",
+      "train Loss: 0.0016\n",
+      "val Loss: 0.0506\n",
       "\n",
-      "Epoch 150/150\n",
-      "train Loss: 0.2953 Acc: 0.8811\n",
-      "val Loss: 0.2982 Acc: 0.8831\n",
+      "Epoch 950/1000\n",
+      "train Loss: 0.0016\n",
+      "val Loss: 0.0515\n",
       "\n",
-      "Training complete in 0m 29s\n",
-      "Best val Acc: 0.886472\n"
+      "Epoch 1000/1000\n",
+      "train Loss: 0.0018\n",
+      "val Loss: 0.0512\n",
+      "\n",
+      "Training complete in 0m 10s\n",
+      "Best val MSE: 0.042776\n"
      ]
     }
    ],
    "source": [
     "# Case 3. CNN_1D (w/o data representation)\n",
-    "model_name = 'CNN_1D'\n",
+    "model_name = 'CNN_1D_rg'\n",
     "model_params = config.model_config[model_name]\n",
     "\n",
-    "data_cls = mc.Classification(model_params)\n",
-    "best_model = data_cls.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
-    "data_cls.save_model(best_model, best_model_path=model_params[\"best_model_path\"])  # 모델 저장"
+    "exp_path = Saver(model_name)\n",
+    "data_reg = mr.Regression(model_params)\n",
+    "best_model = data_reg.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
+    "data_reg.save_model(best_model, best_model_path=os.path.join(exp_path, model_params[\"best_model\"]))  # 모델 저장"
    ]
   },
   {
@@ -357,85 +387,106 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Start training model: LSTM_FCNs\n",
+      "Start training model: LSTM_FCNs_rg\n",
+      "\n",
+      "Epoch 1/1000\n",
+      "train Loss: 0.0531\n",
+      "val Loss: 0.0456\n",
+      "\n",
+      "Epoch 50/1000\n",
+      "train Loss: 0.0156\n",
+      "val Loss: 0.0420\n",
+      "\n",
+      "Epoch 100/1000\n",
+      "train Loss: 0.0070\n",
+      "val Loss: 0.0409\n",
+      "\n",
+      "Epoch 150/1000\n",
+      "train Loss: 0.0034\n",
+      "val Loss: 0.0337\n",
+      "\n",
+      "Epoch 200/1000\n",
+      "train Loss: 0.0036\n",
+      "val Loss: 0.0366\n",
       "\n",
-      "Epoch 1/150\n",
-      "train Loss: 1.4998 Acc: 0.7123\n",
-      "val Loss: 1.3969 Acc: 0.7682\n",
+      "Epoch 250/1000\n",
+      "train Loss: 0.0030\n",
+      "val Loss: 0.0375\n",
       "\n",
-      "Epoch 10/150\n",
-      "train Loss: 0.4978 Acc: 0.9379\n",
-      "val Loss: 0.4642 Acc: 0.9483\n",
+      "Epoch 300/1000\n",
+      "train Loss: 0.0046\n",
+      "val Loss: 0.0367\n",
       "\n",
-      "Epoch 20/150\n",
-      "train Loss: 0.3371 Acc: 0.9515\n",
-      "val Loss: 0.2842 Acc: 0.9579\n",
+      "Epoch 350/1000\n",
+      "train Loss: 0.0025\n",
+      "val Loss: 0.0396\n",
       "\n",
-      "Epoch 30/150\n",
-      "train Loss: 0.1591 Acc: 0.9549\n",
-      "val Loss: 0.1580 Acc: 0.9572\n",
+      "Epoch 400/1000\n",
+      "train Loss: 0.0018\n",
+      "val Loss: 0.0368\n",
       "\n",
-      "Epoch 40/150\n",
-      "train Loss: 0.1276 Acc: 0.9607\n",
-      "val Loss: 0.1133 Acc: 0.9592\n",
+      "Epoch 450/1000\n",
+      "train Loss: 0.0023\n",
+      "val Loss: 0.0320\n",
       "\n",
-      "Epoch 50/150\n",
-      "train Loss: 0.1059 Acc: 0.9626\n",
-      "val Loss: 0.1027 Acc: 0.9613\n",
+      "Epoch 500/1000\n",
+      "train Loss: 0.0029\n",
+      "val Loss: 0.0358\n",
       "\n",
-      "Epoch 60/150\n",
-      "train Loss: 0.0929 Acc: 0.9663\n",
-      "val Loss: 0.0938 Acc: 0.9680\n",
+      "Epoch 550/1000\n",
+      "train Loss: 0.0013\n",
+      "val Loss: 0.0368\n",
       "\n",
-      "Epoch 70/150\n",
-      "train Loss: 0.0834 Acc: 0.9655\n",
-      "val Loss: 0.1311 Acc: 0.9477\n",
+      "Epoch 600/1000\n",
+      "train Loss: 0.0021\n",
+      "val Loss: 0.0345\n",
       "\n",
-      "Epoch 80/150\n",
-      "train Loss: 0.0788 Acc: 0.9679\n",
-      "val Loss: 0.2269 Acc: 0.9048\n",
+      "Epoch 650/1000\n",
+      "train Loss: 0.0013\n",
+      "val Loss: 0.0329\n",
       "\n",
-      "Epoch 90/150\n",
-      "train Loss: 0.0762 Acc: 0.9660\n",
-      "val Loss: 0.0718 Acc: 0.9680\n",
+      "Epoch 700/1000\n",
+      "train Loss: 0.0006\n",
+      "val Loss: 0.0340\n",
       "\n",
-      "Epoch 100/150\n",
-      "train Loss: 0.0748 Acc: 0.9689\n",
-      "val Loss: 0.0922 Acc: 0.9579\n",
+      "Epoch 750/1000\n",
+      "train Loss: 0.0021\n",
+      "val Loss: 0.0347\n",
       "\n",
-      "Epoch 110/150\n",
-      "train Loss: 0.0723 Acc: 0.9685\n",
-      "val Loss: 0.0696 Acc: 0.9680\n",
+      "Epoch 800/1000\n",
+      "train Loss: 0.0028\n",
+      "val Loss: 0.0356\n",
       "\n",
-      "Epoch 120/150\n",
-      "train Loss: 0.0654 Acc: 0.9721\n",
-      "val Loss: 0.3229 Acc: 0.8708\n",
+      "Epoch 850/1000\n",
+      "train Loss: 0.0009\n",
+      "val Loss: 0.0295\n",
       "\n",
-      "Epoch 130/150\n",
-      "train Loss: 0.0666 Acc: 0.9725\n",
-      "val Loss: 0.1340 Acc: 0.9375\n",
+      "Epoch 900/1000\n",
+      "train Loss: 0.0010\n",
+      "val Loss: 0.0311\n",
       "\n",
-      "Epoch 140/150\n",
-      "train Loss: 0.0603 Acc: 0.9728\n",
-      "val Loss: 0.0778 Acc: 0.9667\n",
+      "Epoch 950/1000\n",
+      "train Loss: 0.0006\n",
+      "val Loss: 0.0310\n",
       "\n",
-      "Epoch 150/150\n",
-      "train Loss: 0.0604 Acc: 0.9723\n",
-      "val Loss: 0.1095 Acc: 0.9606\n",
+      "Epoch 1000/1000\n",
+      "train Loss: 0.0008\n",
+      "val Loss: 0.0324\n",
       "\n",
-      "Training complete in 1m 34s\n",
-      "Best val Acc: 0.974847\n"
+      "Training complete in 0m 46s\n",
+      "Best val MSE: 0.026001\n"
      ]
     }
    ],
    "source": [
     "# Case 4. LSTM_FCNs (w/o data representation)\n",
-    "model_name = 'LSTM_FCNs'\n",
+    "model_name = 'LSTM_FCNs_rg'\n",
     "model_params = config.model_config[model_name]\n",
     "\n",
-    "data_cls = mc.Classification(model_params)\n",
-    "best_model = data_cls.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
-    "data_cls.save_model(best_model, best_model_path=model_params[\"best_model_path\"])  # 모델 저장"
+    "exp_path = Saver(model_name)\n",
+    "data_reg = mr.Regression(model_params)\n",
+    "best_model = data_reg.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
+    "data_reg.save_model(best_model, best_model_path=os.path.join(exp_path, model_params[\"best_model\"]))  # 모델 저장"
    ]
   },
   {
@@ -454,14 +505,15 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Save MinMaxScaler in path: ./scaler/minmax_scaler_x_repr.pkl\n"
+      "Save MinMaxScaler in path: ./scaler/minmax_scaler_x_repr.pkl\n",
+      "Save MinMaxScaler in path: ./scaler/minmax_scaler_y_repr.pkl\n"
      ]
     }
    ],
    "source": [
     "# load representation data\n",
     "data_root_dir = './data/'\n",
-    "train_x, train_y, test_x, test_y = utils.load_data(data_root_dir, model_name='FC')  # shape=(num_of_instance, embedding_dim)\n",
+    "train_x, train_y, test_x, test_y = utils.load_data(data_root_dir, model_name='FC_rg')  # shape=(num_of_instance, embedding_dim)\n",
     "\n",
     "# split train data into train/valiation data\n",
     "# train data를 랜덤으로 test_size=split_ratio에 대하여 train/validation set으로 분할\n",
@@ -470,7 +522,9 @@
     "\n",
     "# normalization\n",
     "scaler_x_path = './scaler/minmax_scaler_x_repr.pkl'\n",
-    "train_x, valid_x = utils.get_train_val_data(train_x, valid_x, scaler_x_path)"
+    "scaler_y_path = './scaler/minmax_scaler_y_repr.pkl'\n",
+    "train_x, valid_x = utils.get_train_val_data(train_x, valid_x, scaler_x_path)\n",
+    "train_y, valid_y = utils.get_train_val_data(train_y, valid_y, scaler_y_path)"
    ]
   },
   {
@@ -484,85 +538,106 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Start training model: FC\n",
+      "Start training model: FC_rg\n",
+      "\n",
+      "Epoch 1/1000\n",
+      "train Loss: 0.0522\n",
+      "val Loss: 0.0415\n",
+      "\n",
+      "Epoch 50/1000\n",
+      "train Loss: 0.0433\n",
+      "val Loss: 0.0345\n",
+      "\n",
+      "Epoch 100/1000\n",
+      "train Loss: 0.0416\n",
+      "val Loss: 0.0340\n",
+      "\n",
+      "Epoch 150/1000\n",
+      "train Loss: 0.0416\n",
+      "val Loss: 0.0335\n",
+      "\n",
+      "Epoch 200/1000\n",
+      "train Loss: 0.0392\n",
+      "val Loss: 0.0332\n",
       "\n",
-      "Epoch 1/150\n",
-      "train Loss: 1.7845 Acc: 0.1682\n",
-      "val Loss: 1.7542 Acc: 0.2345\n",
+      "Epoch 250/1000\n",
+      "train Loss: 0.0382\n",
+      "val Loss: 0.0329\n",
       "\n",
-      "Epoch 10/150\n",
-      "train Loss: 1.2704 Acc: 0.6596\n",
-      "val Loss: 1.2287 Acc: 0.6737\n",
+      "Epoch 300/1000\n",
+      "train Loss: 0.0375\n",
+      "val Loss: 0.0327\n",
       "\n",
-      "Epoch 20/150\n",
-      "train Loss: 0.8235 Acc: 0.7495\n",
-      "val Loss: 0.7868 Acc: 0.8389\n",
+      "Epoch 350/1000\n",
+      "train Loss: 0.0368\n",
+      "val Loss: 0.0325\n",
       "\n",
-      "Epoch 30/150\n",
-      "train Loss: 0.6276 Acc: 0.8051\n",
-      "val Loss: 0.5940 Acc: 0.8742\n",
+      "Epoch 400/1000\n",
+      "train Loss: 0.0361\n",
+      "val Loss: 0.0324\n",
       "\n",
-      "Epoch 40/150\n",
-      "train Loss: 0.5202 Acc: 0.8373\n",
-      "val Loss: 0.4832 Acc: 0.8926\n",
+      "Epoch 450/1000\n",
+      "train Loss: 0.0347\n",
+      "val Loss: 0.0323\n",
       "\n",
-      "Epoch 50/150\n",
-      "train Loss: 0.4367 Acc: 0.8708\n",
-      "val Loss: 0.4038 Acc: 0.9028\n",
+      "Epoch 500/1000\n",
+      "train Loss: 0.0344\n",
+      "val Loss: 0.0323\n",
       "\n",
-      "Epoch 60/150\n",
-      "train Loss: 0.3723 Acc: 0.8912\n",
-      "val Loss: 0.3447 Acc: 0.9164\n",
+      "Epoch 550/1000\n",
+      "train Loss: 0.0332\n",
+      "val Loss: 0.0323\n",
       "\n",
-      "Epoch 70/150\n",
-      "train Loss: 0.3287 Acc: 0.8978\n",
-      "val Loss: 0.2992 Acc: 0.9211\n",
+      "Epoch 600/1000\n",
+      "train Loss: 0.0327\n",
+      "val Loss: 0.0323\n",
       "\n",
-      "Epoch 80/150\n",
-      "train Loss: 0.2944 Acc: 0.9112\n",
-      "val Loss: 0.2645 Acc: 0.9259\n",
+      "Epoch 650/1000\n",
+      "train Loss: 0.0309\n",
+      "val Loss: 0.0323\n",
       "\n",
-      "Epoch 90/150\n",
-      "train Loss: 0.2663 Acc: 0.9170\n",
-      "val Loss: 0.2383 Acc: 0.9320\n",
+      "Epoch 700/1000\n",
+      "train Loss: 0.0285\n",
+      "val Loss: 0.0324\n",
       "\n",
-      "Epoch 100/150\n",
-      "train Loss: 0.2400 Acc: 0.9267\n",
-      "val Loss: 0.2173 Acc: 0.9334\n",
+      "Epoch 750/1000\n",
+      "train Loss: 0.0298\n",
+      "val Loss: 0.0325\n",
       "\n",
-      "Epoch 110/150\n",
-      "train Loss: 0.2243 Acc: 0.9286\n",
-      "val Loss: 0.2009 Acc: 0.9368\n",
+      "Epoch 800/1000\n",
+      "train Loss: 0.0284\n",
+      "val Loss: 0.0327\n",
       "\n",
-      "Epoch 120/150\n",
-      "train Loss: 0.2131 Acc: 0.9289\n",
-      "val Loss: 0.1885 Acc: 0.9402\n",
+      "Epoch 850/1000\n",
+      "train Loss: 0.0279\n",
+      "val Loss: 0.0327\n",
       "\n",
-      "Epoch 130/150\n",
-      "train Loss: 0.2020 Acc: 0.9303\n",
-      "val Loss: 0.1785 Acc: 0.9443\n",
+      "Epoch 900/1000\n",
+      "train Loss: 0.0287\n",
+      "val Loss: 0.0328\n",
       "\n",
-      "Epoch 140/150\n",
-      "train Loss: 0.1910 Acc: 0.9335\n",
-      "val Loss: 0.1705 Acc: 0.9456\n",
+      "Epoch 950/1000\n",
+      "train Loss: 0.0297\n",
+      "val Loss: 0.0330\n",
       "\n",
-      "Epoch 150/150\n",
-      "train Loss: 0.1821 Acc: 0.9362\n",
-      "val Loss: 0.1633 Acc: 0.9456\n",
+      "Epoch 1000/1000\n",
+      "train Loss: 0.0247\n",
+      "val Loss: 0.0331\n",
       "\n",
-      "Training complete in 0m 17s\n",
-      "Best val Acc: 0.947655\n"
+      "Training complete in 0m 6s\n",
+      "Best val MSE: 0.032240\n"
      ]
     }
    ],
    "source": [
     "# Case 5. fully-connected layers (w/ data representation)\n",
-    "model_name = 'FC'\n",
+    "model_name = 'FC_rg'\n",
     "model_params = config.model_config[model_name]\n",
     "\n",
-    "data_cls = mc.Classification(model_params)\n",
-    "best_model = data_cls.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
-    "data_cls.save_model(best_model, best_model_path=model_params[\"best_model_path\"])  # 모델 저장"
+    "exp_path = Saver(model_name)\n",
+    "data_reg = mr.Regression(model_params)\n",
+    "best_model = data_reg.train_model(train_x, train_y, valid_x, valid_y)  # 모델 학습\n",
+    "data_reg.save_model(best_model, best_model_path=os.path.join(exp_path, model_params[\"best_model\"]))  # 모델 저장"
    ]
   },
   {
@@ -575,9 +650,9 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "iitp_time_serise",
+   "display_name": "Python 3.7.13 ('hwang')",
    "language": "python",
-   "name": "iitp"
+   "name": "python3"
   },
   "language_info": {
    "codemirror_mode": {
@@ -589,7 +664,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.11"
+   "version": "3.7.13"
   },
   "toc": {
    "base_numbering": 1,
@@ -606,7 +681,7 @@
   },
   "vscode": {
    "interpreter": {
-    "hash": "448579598765c533641c9414ab2d5f7dc2c763c0bd653033bda6f4c5cb841b26"
+    "hash": "4f311b633f277f5e4d199db5715cd5ecdec0742926e2e6ae6b04aa7fd42957fc"
    }
   }
  },
diff --git a/config.py b/config.py
index 52694bf..eeedbf4 100644
--- a/config.py
+++ b/config.py
@@ -1,7 +1,7 @@
 model_config = {
     'LSTM' : {  # Case 1. LSTM model (w/o data representation)
             'model': 'LSTM', 
-            'best_model_path': './ckpt/lstm.pt',  # 학습 완료 모델 저장 경로
+            'best_model': 'lstm.pt',  # 학습 완료 모델 저장 경로
             'parameter': {
                 'input_size': 9,  # 데이터의 변수 개수, int
                 'num_classes': 6,  # 분류할 class 개수, int
@@ -17,7 +17,7 @@
     },
     'GRU' : {  # Case 2. GRU model (w/o data representation)
         'model': 'GRU',
-        'best_model_path': './ckpt/gru.pt',  # 학습 완료 모델 저장 경로
+        'best_model': 'gru.pt',  # 학습 완료 모델 저장 경로
         'parameter': {
             'input_size': 9,  # 데이터의 변수 개수, int
             'num_classes': 6,  # 분류할 class 개수, int
@@ -33,7 +33,7 @@
     },
     'CNN_1D' : {  # Case 3. CNN_1D model (w/o data representation)
         'model': 'CNN_1D', 
-        'best_model_path': './ckpt/cnn_1d.pt',  # 학습 완료 모델 저장 경로
+        'best_model': 'cnn_1d.pt',  # 학습 완료 모델 저장 경로
         'parameter': {
             'input_size': 9,  # 데이터의 변수 개수, int
             'num_classes': 6,  # 분류할 class 개수, int
@@ -51,7 +51,7 @@
     },
     'LSTM_FCNs' : {  # Case 4. LSTM_FCNs model (w/o data representation)
         'model': 'LSTM_FCNs', 
-        'best_model_path': './ckpt/lstm_fcn.pt',  # 학습 완료 모델 저장 경로
+        'best_model': 'lstm_fcn.pt',  # 학습 완료 모델 저장 경로
         'parameter': {
             'input_size': 9,  # 데이터의 변수 개수, int
             'num_classes': 6,  # 분류할 class 개수, int
@@ -66,7 +66,7 @@
     },
     'FC' : {  # Case 5. fully-connected layers (w/ data representation)
         'model': 'FC', 
-        "best_model_path": './ckpt/fc.pt',  # 학습 완료 모델 저장 경로
+        "best_model": 'fc.pt',  # 학습 완료 모델 저장 경로
         'parameter': {
             'input_size': 64,  # 데이터의 변수 개수(representation 차원), int
             'num_classes': 6,  # 분류할 class 개수, int