diff --git a/机器学习竞赛实战_优胜解决方案/特征工程建模/模型解释.ipynb b/机器学习竞赛实战_优胜解决方案/特征工程建模/模型解释.ipynb
index 66c20d6..8c47cab 100644
--- a/机器学习竞赛实战_优胜解决方案/特征工程建模/模型解释.ipynb
+++ b/机器学习竞赛实战_优胜解决方案/特征工程建模/模型解释.ipynb
@@ -9,7 +9,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -21,7 +21,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -226,7 +226,7 @@
        "[5 rows x 27 columns]"
       ]
      },
-     "execution_count": 3,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -251,7 +251,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -265,7 +265,7 @@
        "Name: Man of the Match, dtype: bool"
       ]
      },
-     "execution_count": 4,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -277,7 +277,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
@@ -461,7 +461,7 @@
        "4             0  "
       ]
      },
-     "execution_count": 5,
+     "execution_count": 4,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -474,7 +474,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [
     {
@@ -493,7 +493,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -795,7 +795,7 @@
        "<IPython.core.display.HTML object>"
       ]
      },
-     "execution_count": 15,
+     "execution_count": 6,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -817,8 +817,56 @@
     "\n",
     "特征重要性展示了每个特征发挥的作用情况，partial dependence plots 可以展示一个特征怎么影响预测结果\n",
     "\n",
-    "模型建立完成后进行使用，概述如下："
+    "模型建立完成后进行使用，概述如下：\n",
+    "<ul>\n",
+    "    <li>首先选中一个样本数据，此时想观察Ball Possession列对结果的影响。\n",
+    "    <li>保证其它特征列不变，改变当前观察列的值，例如悬着40%，50%，60%分别进行预测，得到各自的结果。\n",
+    "    <li>对比结果就能知道当前列（Ball Possession）对结果的影响情况。"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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\n",
+      "text/plain": [
+       "<Figure size 1080x684 with 2 Axes>"
+      ]
+     },
+     "metadata": {
+      "needs_background": "light"
+     },
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "from pdpbox import pdp, get_dataset, info_plots  # pip install pdpbox\n",
+    "pdp_goals = pdp.pdp_isolate(model=my_model,\n",
+    "                            dataset=val_X, \n",
+    "                            model_features=feature_names,\n",
+    "                            feature='Goal Scored')  # 观察的特征\n",
+    "\n",
+    "pdp.pdp_plot(pdp_goals, 'Goal Scored')\n",
+    "plt.show()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "y轴表示预测结果的变化对比于基本模型，由于在观察时不可能只看一个样本数据，肯定要选择多个样本数据，蓝色区域表示的是置信度"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {