From 6ca48876923f6ca87e1c9804dc9d122c60113ef1 Mon Sep 17 00:00:00 2001
From: benjas <909336740@qq.com>
Date: Fri, 19 Feb 2021 00:16:32 +0800
Subject: [PATCH] Add. Summary

---
 .../NLP处理实例.ipynb                        | 16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)

diff --git a/机器学习竞赛实战_优胜解决方案/机器学习实战小项目/文本特征处理方法对比/NLP处理实例.ipynb b/机器学习竞赛实战_优胜解决方案/机器学习实战小项目/文本特征处理方法对比/NLP处理实例.ipynb
index 74aa39a..178f790 100644
--- a/机器学习竞赛实战_优胜解决方案/机器学习实战小项目/文本特征处理方法对比/NLP处理实例.ipynb
+++ b/机器学习竞赛实战_优胜解决方案/机器学习实战小项目/文本特征处理方法对比/NLP处理实例.ipynb
@@ -1380,7 +1380,7 @@
     "import gensim\n",
     "\n",
     "# 读取预训练模型\n",
-    "word2vec_path = \"GoogleNews-vectors-negative300.bin\"\n",
+    "word2vec_path = \"GoogleNews-vectors-negative300.bin\"  # 下载地址：https://s3.amazonaws.com/dl4j-distribution/GoogleNews-vectors-negative300.bin.gz\n",
     "word2vec = gensim.models.KeyedVectors.load_word2vec_format(word2vec_path, binary=True)"
    ]
   },
@@ -76194,11 +76194,17 @@
    ]
   },
   {
-   "cell_type": "code",
-   "execution_count": null,
+   "cell_type": "markdown",
    "metadata": {},
-   "outputs": [],
-   "source": []
+   "source": [
+    "### 总结\n",
+    "\n",
+    "* 文本数据同样需要预处理\n",
+    "* 预处理完对文本做Embedding\n",
+    "* 再利用模型训练及预测\n",
+    "\n",
+    "以目前来看神经网络比传统模型效果更好，但实际场景中往往是传统可解释的模型更优，我们知道除开技术更重要的是应用到实际场景中，从而需要告诉业务，解释给业务，这样才能发挥更大的效能。另外，可能"
+   ]
   }
  ],
  "metadata": {