diff --git a/NN_Training.ipynb b/NN_Training.ipynb
index 72239d3..08e883e 100644
--- a/NN_Training.ipynb
+++ b/NN_Training.ipynb
@@ -28,7 +28,22 @@
    "metadata": {
     "id": "Fyk7fzYbPQFL"
    },
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "2025-05-18 20:35:28.442029: I external/local_tsl/tsl/cuda/cudart_stub.cc:31] Could not find cuda drivers on your machine, GPU will not be used.\n",
+      "2025-05-18 20:35:28.476823: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered\n",
+      "2025-05-18 20:35:28.476876: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered\n",
+      "2025-05-18 20:35:28.478169: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered\n",
+      "2025-05-18 20:35:28.484633: I external/local_tsl/tsl/cuda/cudart_stub.cc:31] Could not find cuda drivers on your machine, GPU will not be used.\n",
+      "2025-05-18 20:35:28.485718: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations.\n",
+      "To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
+      "2025-05-18 20:35:29.551411: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Could not find TensorRT\n"
+     ]
+    }
+   ],
    "source": [
     "# imports\n",
     "import numpy as np\n",
@@ -53,7 +68,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -79,13 +94,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [],
    "source": [
     "# train-test split for model evaluation\n",
     "X_train, X_test, y_train, y_test = train_test_split(\n",
-    "    X, y, train_size=0.8, shuffle=True\n",
+    "    X, y, train_size=0.8, random_state=42, shuffle=True\n",
     ")"
    ]
   },
@@ -98,14 +113,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "/var/folders/d4/trj3dssd6s3g8kxvjmczz11w0000gn/T/ipykernel_75912/4103796048.py:1: DeprecationWarning: Please import `LinearNDInterpolator` from the `scipy.interpolate` namespace; the `scipy.interpolate.interpnd` namespace is deprecated and will be removed in SciPy 2.0.0.\n",
+      "/tmp/ipykernel_398158/4103796048.py:1: DeprecationWarning: Please import `LinearNDInterpolator` from the `scipy.interpolate` namespace; the `scipy.interpolate.interpnd` namespace is deprecated and will be removed in SciPy 2.0.0.\n",
       "  interpolant_avg_diff = pickle.load(\n"
      ]
     },
@@ -113,11 +128,11 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Predictions =  [[2472.26461622]\n",
-      " [1701.04342628]\n",
-      " [ 682.2876416 ]\n",
-      " [1760.64694548]\n",
-      " [1204.88980962]]\n"
+      "Predictions =  [[1540.55086271]\n",
+      " [2460.84864787]\n",
+      " [ 456.17321004]\n",
+      " [1149.23911813]\n",
+      " [1503.92608606]]\n"
      ]
     }
    ],
@@ -141,7 +156,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -156,7 +171,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -183,48 +198,52 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
     "# Define the normalization function of the output\n",
-    "def normalize_output(data):\n",
-    "    data_max = np.max(data)\n",
-    "    data_min = np.min(data)\n",
-    "    return (data - data_min) / (data_max - data_min)"
+    "# Define the normalization function of the output\n",
+    "def normalize_output(data, data_max=None, data_min=None):\n",
+    "    \"\"\"Normalizes the output data to a [0, 1] range.\"\"\"\n",
+    "    if data_max is None:\n",
+    "        data_max = np.max(data)\n",
+    "    if data_min is None:\n",
+    "        data_min = np.min(data)\n",
+    "    return (data - data_min) / (data_max - data_min)\n"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Model: \"sequential_1\"\n",
+      "Model: \"sequential\"\n",
       "_________________________________________________________________\n",
       " Layer (type)                Output Shape              Param #   \n",
       "=================================================================\n",
-      " normalization_1 (Normaliza  (None, 6)                 13        \n",
-      " tion)                                                           \n",
+      " normalization (Normalizati  (None, 6)                 13        \n",
+      " on)                                                             \n",
       "                                                                 \n",
-      " dense_8 (Dense)             (None, 120)               840       \n",
+      " dense (Dense)               (None, 120)               840       \n",
       "                                                                 \n",
-      " dense_9 (Dense)             (None, 108)               13068     \n",
+      " dense_1 (Dense)             (None, 108)               13068     \n",
       "                                                                 \n",
-      " dense_10 (Dense)            (None, 96)                10464     \n",
+      " dense_2 (Dense)             (None, 96)                10464     \n",
       "                                                                 \n",
-      " dense_11 (Dense)            (None, 84)                8148      \n",
+      " dense_3 (Dense)             (None, 84)                8148      \n",
       "                                                                 \n",
-      " dense_12 (Dense)            (None, 72)                6120      \n",
+      " dense_4 (Dense)             (None, 72)                6120      \n",
       "                                                                 \n",
-      " dense_13 (Dense)            (None, 60)                4380      \n",
+      " dense_5 (Dense)             (None, 60)                4380      \n",
       "                                                                 \n",
-      " dense_14 (Dense)            (None, 48)                2928      \n",
+      " dense_6 (Dense)             (None, 48)                2928      \n",
       "                                                                 \n",
-      " dense_15 (Dense)            (None, 1)                 49        \n",
+      " dense_7 (Dense)             (None, 1)                 49        \n",
       "                                                                 \n",
       "=================================================================\n",
       "Total params: 46010 (179.73 KB)\n",
@@ -264,7 +283,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 9,
    "metadata": {
     "id": "YRwvbbr0XIbf"
    },
@@ -284,7 +303,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [
     {
@@ -292,216 +311,218 @@
      "output_type": "stream",
      "text": [
       "Epoch 1/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 0.0045 - mean_squared_error: 0.0045\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 0.0045 - mean_squared_error: 0.0045\n",
       "Epoch 2/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.1973e-04 - mean_squared_error: 4.1973e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.8090e-04 - mean_squared_error: 4.8090e-04\n",
       "Epoch 3/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.5104e-04 - mean_squared_error: 2.5104e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.0310e-04 - mean_squared_error: 3.0310e-04\n",
       "Epoch 4/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.1135e-04 - mean_squared_error: 2.1135e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.3339e-04 - mean_squared_error: 2.3339e-04\n",
       "Epoch 5/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.9093e-04 - mean_squared_error: 1.9093e-04 - val_loss: 2.1363e-04 - val_mean_squared_error: 2.1363e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.0115e-04 - mean_squared_error: 2.0115e-04 - val_loss: 1.9130e-04 - val_mean_squared_error: 1.9130e-04\n",
       "Epoch 6/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.7413e-04 - mean_squared_error: 1.7413e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.6831e-04 - mean_squared_error: 1.6831e-04\n",
       "Epoch 7/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.6051e-04 - mean_squared_error: 1.6051e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.9413e-04 - mean_squared_error: 1.9413e-04\n",
       "Epoch 8/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.7333e-04 - mean_squared_error: 1.7333e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.4554e-04 - mean_squared_error: 1.4554e-04\n",
       "Epoch 9/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.4864e-04 - mean_squared_error: 1.4864e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.5658e-04 - mean_squared_error: 1.5658e-04\n",
       "Epoch 10/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.5507e-04 - mean_squared_error: 1.5507e-04 - val_loss: 1.4615e-04 - val_mean_squared_error: 1.4615e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 1.3820e-04 - mean_squared_error: 1.3820e-04 - val_loss: 1.3510e-04 - val_mean_squared_error: 1.3510e-04\n",
       "Epoch 11/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.2418e-04 - mean_squared_error: 1.2418e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.2747e-04 - mean_squared_error: 1.2747e-04\n",
       "Epoch 12/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.0832e-04 - mean_squared_error: 1.0832e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.2902e-04 - mean_squared_error: 1.2902e-04\n",
       "Epoch 13/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.0593e-04 - mean_squared_error: 1.0593e-04\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.1640e-04 - mean_squared_error: 1.1640e-04\n",
       "Epoch 14/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 9.7592e-05 - mean_squared_error: 9.7592e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.1147e-04 - mean_squared_error: 1.1147e-04\n",
       "Epoch 15/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.1264e-04 - mean_squared_error: 1.1264e-04 - val_loss: 2.1124e-04 - val_mean_squared_error: 2.1124e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 1.0179e-04 - mean_squared_error: 1.0179e-04 - val_loss: 1.1728e-04 - val_mean_squared_error: 1.1728e-04\n",
       "Epoch 16/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 9.2550e-05 - mean_squared_error: 9.2550e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 1.0370e-04 - mean_squared_error: 1.0370e-04\n",
       "Epoch 17/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 8.6084e-05 - mean_squared_error: 8.6084e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 9.9976e-05 - mean_squared_error: 9.9976e-05\n",
       "Epoch 18/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 8.4333e-05 - mean_squared_error: 8.4333e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 8.9677e-05 - mean_squared_error: 8.9677e-05\n",
       "Epoch 19/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 7.7898e-05 - mean_squared_error: 7.7898e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 8.3252e-05 - mean_squared_error: 8.3252e-05\n",
       "Epoch 20/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 7.3489e-05 - mean_squared_error: 7.3489e-05 - val_loss: 1.4588e-04 - val_mean_squared_error: 1.4588e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 9.0918e-05 - mean_squared_error: 9.0918e-05 - val_loss: 7.6746e-05 - val_mean_squared_error: 7.6746e-05\n",
       "Epoch 21/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 7.9932e-05 - mean_squared_error: 7.9932e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 7.5562e-05 - mean_squared_error: 7.5562e-05\n",
       "Epoch 22/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 7.5347e-05 - mean_squared_error: 7.5347e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 7.5469e-05 - mean_squared_error: 7.5469e-05\n",
       "Epoch 23/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 6.1833e-05 - mean_squared_error: 6.1833e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 6.7959e-05 - mean_squared_error: 6.7959e-05\n",
       "Epoch 24/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 6.4398e-05 - mean_squared_error: 6.4398e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 6.6471e-05 - mean_squared_error: 6.6471e-05\n",
       "Epoch 25/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 6.5495e-05 - mean_squared_error: 6.5495e-05 - val_loss: 9.7698e-05 - val_mean_squared_error: 9.7698e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 7.0012e-05 - mean_squared_error: 7.0012e-05 - val_loss: 6.6933e-05 - val_mean_squared_error: 6.6933e-05\n",
       "Epoch 26/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 5.8114e-05 - mean_squared_error: 5.8114e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 6.2621e-05 - mean_squared_error: 6.2621e-05\n",
       "Epoch 27/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 5.5296e-05 - mean_squared_error: 5.5296e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 5.7108e-05 - mean_squared_error: 5.7108e-05\n",
       "Epoch 28/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 5.5693e-05 - mean_squared_error: 5.5693e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 5.6734e-05 - mean_squared_error: 5.6734e-05\n",
       "Epoch 29/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 6.0006e-05 - mean_squared_error: 6.0006e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 5.8363e-05 - mean_squared_error: 5.8363e-05\n",
       "Epoch 30/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 5.2514e-05 - mean_squared_error: 5.2514e-05 - val_loss: 1.3176e-04 - val_mean_squared_error: 1.3176e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 5.7048e-05 - mean_squared_error: 5.7048e-05 - val_loss: 5.4878e-05 - val_mean_squared_error: 5.4878e-05\n",
       "Epoch 31/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 5.0566e-05 - mean_squared_error: 5.0566e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 5.5206e-05 - mean_squared_error: 5.5206e-05\n",
       "Epoch 32/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.7183e-05 - mean_squared_error: 4.7183e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 5.5958e-05 - mean_squared_error: 5.5958e-05\n",
       "Epoch 33/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.9724e-05 - mean_squared_error: 4.9724e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.8364e-05 - mean_squared_error: 4.8364e-05\n",
       "Epoch 34/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.6217e-05 - mean_squared_error: 4.6217e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 5.0521e-05 - mean_squared_error: 5.0521e-05\n",
       "Epoch 35/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.4077e-05 - mean_squared_error: 4.4077e-05 - val_loss: 1.0123e-04 - val_mean_squared_error: 1.0123e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 4.7853e-05 - mean_squared_error: 4.7853e-05 - val_loss: 5.4748e-05 - val_mean_squared_error: 5.4748e-05\n",
       "Epoch 36/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.5058e-05 - mean_squared_error: 4.5058e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.7025e-05 - mean_squared_error: 4.7025e-05\n",
       "Epoch 37/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.3521e-05 - mean_squared_error: 4.3521e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.3510e-05 - mean_squared_error: 4.3510e-05\n",
       "Epoch 38/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.3273e-05 - mean_squared_error: 4.3273e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.3774e-05 - mean_squared_error: 4.3774e-05\n",
       "Epoch 39/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.2109e-05 - mean_squared_error: 4.2109e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.5029e-05 - mean_squared_error: 4.5029e-05\n",
       "Epoch 40/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.6930e-05 - mean_squared_error: 3.6930e-05 - val_loss: 9.2651e-05 - val_mean_squared_error: 9.2651e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 4.4976e-05 - mean_squared_error: 4.4976e-05 - val_loss: 4.8876e-05 - val_mean_squared_error: 4.8876e-05\n",
       "Epoch 41/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.8375e-05 - mean_squared_error: 3.8375e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 4.1319e-05 - mean_squared_error: 4.1319e-05\n",
       "Epoch 42/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.0656e-05 - mean_squared_error: 4.0656e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 4.1494e-05 - mean_squared_error: 4.1494e-05\n",
       "Epoch 43/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.8336e-05 - mean_squared_error: 3.8336e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.9689e-05 - mean_squared_error: 3.9689e-05\n",
       "Epoch 44/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.9249e-05 - mean_squared_error: 3.9249e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 4.1729e-05 - mean_squared_error: 4.1729e-05\n",
       "Epoch 45/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 4.0144e-05 - mean_squared_error: 4.0144e-05 - val_loss: 9.5994e-05 - val_mean_squared_error: 9.5994e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.8982e-05 - mean_squared_error: 3.8982e-05 - val_loss: 7.2035e-05 - val_mean_squared_error: 7.2035e-05\n",
       "Epoch 46/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.8839e-05 - mean_squared_error: 3.8839e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.7213e-05 - mean_squared_error: 3.7213e-05\n",
       "Epoch 47/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.3104e-05 - mean_squared_error: 3.3104e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.9416e-05 - mean_squared_error: 3.9416e-05\n",
       "Epoch 48/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.3755e-05 - mean_squared_error: 3.3755e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.6523e-05 - mean_squared_error: 3.6523e-05\n",
       "Epoch 49/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.5519e-05 - mean_squared_error: 3.5519e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.4498e-05 - mean_squared_error: 3.4498e-05\n",
       "Epoch 50/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.3671e-05 - mean_squared_error: 3.3671e-05 - val_loss: 1.3716e-04 - val_mean_squared_error: 1.3716e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.6180e-05 - mean_squared_error: 3.6180e-05 - val_loss: 4.3387e-05 - val_mean_squared_error: 4.3387e-05\n",
       "Epoch 51/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.6045e-05 - mean_squared_error: 3.6045e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.3777e-05 - mean_squared_error: 3.3777e-05\n",
       "Epoch 52/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.3578e-05 - mean_squared_error: 3.3578e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.3207e-05 - mean_squared_error: 3.3207e-05\n",
       "Epoch 53/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.0851e-05 - mean_squared_error: 3.0851e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.6231e-05 - mean_squared_error: 3.6231e-05\n",
       "Epoch 54/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.1669e-05 - mean_squared_error: 3.1669e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.2489e-05 - mean_squared_error: 3.2489e-05\n",
       "Epoch 55/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.1917e-05 - mean_squared_error: 3.1917e-05 - val_loss: 1.1290e-04 - val_mean_squared_error: 1.1290e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.4902e-05 - mean_squared_error: 3.4902e-05 - val_loss: 6.1230e-05 - val_mean_squared_error: 6.1230e-05\n",
       "Epoch 56/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.4160e-05 - mean_squared_error: 3.4160e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.4406e-05 - mean_squared_error: 3.4406e-05\n",
       "Epoch 57/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.2770e-05 - mean_squared_error: 3.2770e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.2576e-05 - mean_squared_error: 3.2576e-05\n",
       "Epoch 58/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.7974e-05 - mean_squared_error: 2.7974e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.1893e-05 - mean_squared_error: 3.1893e-05\n",
       "Epoch 59/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.9894e-05 - mean_squared_error: 2.9894e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.1560e-05 - mean_squared_error: 3.1560e-05\n",
       "Epoch 60/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.1467e-05 - mean_squared_error: 3.1467e-05 - val_loss: 1.1368e-04 - val_mean_squared_error: 1.1368e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.2778e-05 - mean_squared_error: 3.2778e-05 - val_loss: 3.6462e-05 - val_mean_squared_error: 3.6462e-05\n",
       "Epoch 61/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.8880e-05 - mean_squared_error: 2.8880e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 3.0391e-05 - mean_squared_error: 3.0391e-05\n",
       "Epoch 62/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.4254e-05 - mean_squared_error: 3.4254e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.8066e-05 - mean_squared_error: 2.8066e-05\n",
       "Epoch 63/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.6570e-05 - mean_squared_error: 2.6570e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.8828e-05 - mean_squared_error: 2.8828e-05\n",
       "Epoch 64/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.7210e-05 - mean_squared_error: 2.7210e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.0239e-05 - mean_squared_error: 3.0239e-05\n",
       "Epoch 65/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.7584e-05 - mean_squared_error: 2.7584e-05 - val_loss: 1.8097e-04 - val_mean_squared_error: 1.8097e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.7537e-05 - mean_squared_error: 2.7537e-05 - val_loss: 3.6516e-05 - val_mean_squared_error: 3.6516e-05\n",
       "Epoch 66/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 3.5441e-05 - mean_squared_error: 3.5441e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.6745e-05 - mean_squared_error: 2.6745e-05\n",
       "Epoch 67/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.8004e-05 - mean_squared_error: 2.8004e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.0606e-05 - mean_squared_error: 3.0606e-05\n",
       "Epoch 68/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.6565e-05 - mean_squared_error: 2.6565e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.8006e-05 - mean_squared_error: 2.8006e-05\n",
       "Epoch 69/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.5695e-05 - mean_squared_error: 2.5695e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 3.1340e-05 - mean_squared_error: 3.1340e-05\n",
       "Epoch 70/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.4909e-05 - mean_squared_error: 2.4909e-05 - val_loss: 1.6851e-04 - val_mean_squared_error: 1.6851e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.8085e-05 - mean_squared_error: 2.8085e-05 - val_loss: 3.0832e-05 - val_mean_squared_error: 3.0832e-05\n",
       "Epoch 71/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.5246e-05 - mean_squared_error: 2.5246e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.4840e-05 - mean_squared_error: 2.4840e-05\n",
       "Epoch 72/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.7716e-05 - mean_squared_error: 2.7716e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.8356e-05 - mean_squared_error: 2.8356e-05\n",
       "Epoch 73/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.6864e-05 - mean_squared_error: 2.6864e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.6042e-05 - mean_squared_error: 2.6042e-05\n",
       "Epoch 74/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.7754e-05 - mean_squared_error: 2.7754e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.8381e-05 - mean_squared_error: 2.8381e-05\n",
       "Epoch 75/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.5588e-05 - mean_squared_error: 2.5588e-05 - val_loss: 1.2734e-04 - val_mean_squared_error: 1.2734e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.8149e-05 - mean_squared_error: 2.8149e-05 - val_loss: 3.3545e-05 - val_mean_squared_error: 3.3545e-05\n",
       "Epoch 76/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.4143e-05 - mean_squared_error: 2.4143e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.7495e-05 - mean_squared_error: 2.7495e-05\n",
       "Epoch 77/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.3177e-05 - mean_squared_error: 2.3177e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.4941e-05 - mean_squared_error: 2.4941e-05\n",
       "Epoch 78/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.3311e-05 - mean_squared_error: 2.3311e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.5548e-05 - mean_squared_error: 2.5548e-05\n",
       "Epoch 79/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.4370e-05 - mean_squared_error: 2.4370e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.6092e-05 - mean_squared_error: 2.6092e-05\n",
       "Epoch 80/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.4295e-05 - mean_squared_error: 2.4295e-05 - val_loss: 9.3401e-05 - val_mean_squared_error: 9.3401e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.3585e-05 - mean_squared_error: 2.3585e-05 - val_loss: 4.2193e-05 - val_mean_squared_error: 4.2193e-05\n",
       "Epoch 81/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.5748e-05 - mean_squared_error: 2.5748e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.4953e-05 - mean_squared_error: 2.4953e-05\n",
       "Epoch 82/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.3221e-05 - mean_squared_error: 2.3221e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.2610e-05 - mean_squared_error: 2.2610e-05\n",
       "Epoch 83/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2634e-05 - mean_squared_error: 2.2634e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.3686e-05 - mean_squared_error: 2.3686e-05\n",
       "Epoch 84/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.3076e-05 - mean_squared_error: 2.3076e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.5058e-05 - mean_squared_error: 2.5058e-05\n",
       "Epoch 85/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.4865e-05 - mean_squared_error: 2.4865e-05 - val_loss: 1.3313e-04 - val_mean_squared_error: 1.3313e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.4400e-05 - mean_squared_error: 2.4400e-05 - val_loss: 3.1305e-05 - val_mean_squared_error: 3.1305e-05\n",
       "Epoch 86/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2553e-05 - mean_squared_error: 2.2553e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.3477e-05 - mean_squared_error: 2.3477e-05\n",
       "Epoch 87/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2353e-05 - mean_squared_error: 2.2353e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.3354e-05 - mean_squared_error: 2.3354e-05\n",
       "Epoch 88/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2928e-05 - mean_squared_error: 2.2928e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.9633e-05 - mean_squared_error: 2.9633e-05\n",
       "Epoch 89/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2123e-05 - mean_squared_error: 2.2123e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.2263e-05 - mean_squared_error: 2.2263e-05\n",
       "Epoch 90/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2759e-05 - mean_squared_error: 2.2759e-05 - val_loss: 1.3881e-04 - val_mean_squared_error: 1.3881e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.3452e-05 - mean_squared_error: 2.3452e-05 - val_loss: 2.9688e-05 - val_mean_squared_error: 2.9688e-05\n",
       "Epoch 91/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2808e-05 - mean_squared_error: 2.2808e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.2283e-05 - mean_squared_error: 2.2283e-05\n",
       "Epoch 92/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.1738e-05 - mean_squared_error: 2.1738e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.1718e-05 - mean_squared_error: 2.1718e-05\n",
       "Epoch 93/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.1839e-05 - mean_squared_error: 2.1839e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.2278e-05 - mean_squared_error: 2.2278e-05\n",
       "Epoch 94/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.2090e-05 - mean_squared_error: 2.2090e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.1102e-05 - mean_squared_error: 2.1102e-05\n",
       "Epoch 95/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.9664e-05 - mean_squared_error: 1.9664e-05 - val_loss: 1.3487e-04 - val_mean_squared_error: 1.3487e-04\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.2953e-05 - mean_squared_error: 2.2953e-05 - val_loss: 3.0545e-05 - val_mean_squared_error: 3.0545e-05\n",
       "Epoch 96/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.1539e-05 - mean_squared_error: 2.1539e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.5788e-05 - mean_squared_error: 2.5788e-05\n",
       "Epoch 97/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.5831e-05 - mean_squared_error: 2.5831e-05\n",
+      "687/687 [==============================] - 2s 2ms/step - loss: 2.2718e-05 - mean_squared_error: 2.2718e-05\n",
       "Epoch 98/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 1.9564e-05 - mean_squared_error: 1.9564e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.1534e-05 - mean_squared_error: 2.1534e-05\n",
       "Epoch 99/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.0803e-05 - mean_squared_error: 2.0803e-05\n",
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.0597e-05 - mean_squared_error: 2.0597e-05\n",
       "Epoch 100/100\n",
-      "687/687 [==============================] - 1s 1ms/step - loss: 2.0234e-05 - mean_squared_error: 2.0234e-05 - val_loss: 1.3623e-04 - val_mean_squared_error: 1.3623e-04\n"
+      "687/687 [==============================] - 2s 3ms/step - loss: 2.3118e-05 - mean_squared_error: 2.3118e-05 - val_loss: 3.1274e-05 - val_mean_squared_error: 3.1274e-05\n"
      ]
     }
    ],
    "source": [
+    "y_max_train = y_train.max()\n",
+    "y_min_train = y_train.min()\n",
     "history = nn.fit(\n",
     "    X_train, normalize_output(y_train),\n",
     "    epochs=num_epochs,\n",
     "    callbacks=[callback],\n",
     "    # sample_weight=sample_weights,\n",
     "    batch_size=32,\n",
-    "    validation_data=(X_test,  normalize_output(y_test)),\n",
+    "    validation_data=(X_test,  normalize_output(y_test, y_max_train, y_min_train)),\n",
     "    validation_freq=5,   # Evaluate every 5 epochs\n",
     "    verbose=1\n",
     ")"
@@ -509,12 +530,12 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
      "data": {
-      "image/png": "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",
+      "image/png": "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",
       "text/plain": [
        "<Figure size 640x480 with 1 Axes>"
       ]
@@ -547,29 +568,29 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "859/859 - 0s - loss: 1.2431e-04 - mean_squared_error: 1.2431e-04 - 315ms/epoch - 367us/step\n"
+      "859/859 - 1s - loss: 2.1916e-05 - mean_squared_error: 2.1916e-05 - 1s/epoch - 1ms/step\n"
      ]
     },
     {
      "data": {
       "text/plain": [
-       "[0.00012431027425918728, 0.00012431027425918728]"
+       "[2.191568273701705e-05, 2.191568273701705e-05]"
       ]
      },
-     "execution_count": 14,
+     "execution_count": 12,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
-    "nn.evaluate(X,  normalize_output(y), verbose=2)"
+    "nn.evaluate(X,  normalize_output(y, y_max_train, y_min_train), verbose=2)"
    ]
   },
   {
@@ -581,17 +602,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "859/859 [==============================] - 0s 392us/step\n",
-      "The mean abs diff between y and y_pred is 43.452346028208204.\n",
-      "The max diff between y and y_pred is 296.6641874277152.\n",
-      "The min diff between y and y_pred is -193.34191910188383.\n"
+      "859/859 [==============================] - 1s 1ms/step\n",
+      "The mean abs diff between y and y_pred is 15.792049256049639.\n",
+      "The max diff between y and y_pred is 215.51506370501784.\n",
+      "The min diff between y and y_pred is -313.25722740240917.\n"
      ]
     }
    ],
@@ -599,7 +620,7 @@
     "output_variable = \"avg_diff\"\n",
     "predictions_normed = nn.predict(X)\n",
     "# Rescale the predictions\n",
-    "predictions = predictions_normed * (y.max() - y.min()) + y.min()\n",
+    "predictions = predictions_normed * (y_max_train - y_min_train) + y_min_train\n",
     "y_pred = predictions\n",
     "y_diff = y - y_pred\n",
     "print(\n",
@@ -611,15 +632,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 29,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Mean:  [[54.73748     8.35946     8.381766    1.8990896   1.8962128   0.53320134]]\n",
-      "Std dev:  [[5.95917    4.787876   4.805051   1.4172095  1.4169149  0.39561728]]\n"
+      "Mean:  [[54.774956   8.350455   8.384443   1.8986119  1.9025313  0.5333266]]\n",
+      "Std dev:  [[5.9505095  4.8013268  4.798767   1.4162183  1.4184183  0.39660984]]\n"
      ]
     }
    ],
@@ -642,17 +663,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "15/15 [==============================] - 0s 512us/step\n",
-      "The mean abs diff between y_book_pred and y_book is 39.115684254982874.\n",
-      "The max diff between y_book_pred and y_book is 38.33527331989353.\n",
-      "The min diff between y_book_pred and y_book is -130.4563545554597.\n"
+      "15/15 [==============================] - 0s 1ms/step\n",
+      "The mean abs diff between y_book_pred and y_book is 19.13597747169837.\n",
+      "The max diff between y_book_pred and y_book is 80.8457985770549.\n",
+      "The min diff between y_book_pred and y_book is -76.27483755019.\n"
      ]
     }
    ],
@@ -663,7 +684,7 @@
     "\n",
     "# predict using the model\n",
     "y_book_pred_norm = nn.predict(X_book)\n",
-    "y_book_pred = y_book_pred_norm * (y.max() - y.min()) + y.min()\n",
+    "y_book_pred = y_book_pred_norm * (y_max_train - y_min_train) + y_min_train\n",
     "y_book_diff = y_book - y_book_pred\n",
     "print(\n",
     "    f\"The mean abs diff between y_book_pred and y_book \" +\n",
@@ -686,7 +707,7 @@
    "provenance": []
   },
   "kernelspec": {
-   "display_name": "dynviz-dev",
+   "display_name": "Python 3 (ipykernel)",
    "language": "python",
    "name": "python3"
   },
@@ -704,5 +725,5 @@
   }
  },
  "nbformat": 4,
- "nbformat_minor": 0
+ "nbformat_minor": 4
 }