Merge pull request #139 from numerai/anson/ensemble

Anson/ensemble
numerai · Aug 3, 2023 · bb72ff4 · bb72ff4
2 parents 2c0ba0d + 21c79d4
commit bb72ff4
Show file tree

Hide file tree

Showing 3 changed files with 1,746 additions and 183 deletions.
diff --git a/feature_neutralization.ipynb b/feature_neutralization.ipynb
diff --git a/hello_numerai.ipynb b/hello_numerai.ipynb
@@ -14,7 +14,10 @@
    "source": [
     "Hello and welcome to the Numerai Data Science Tournament!\n",
     "\n",
-    "In this notebook we will explore the basics of the Numerai dataset and build our first machine learning model."
+    "In this notebook we will\n",
+    "1. Download and explore the dataset\n",
+    "2. Train our first machine learning model\n",
+    "3. Upload our model to start making live submissions"
    ]
   },
   {
@@ -471,7 +474,7 @@
    "source": [
     "The primary scoring metric in Numerai is called `numerai_corr` or just `corr`, which is a Numerai specific variant of the Pearson Correlation Coefficient. This scoring metric is designed to \"align incentives\" between model scoring and hedge fund performance. In other words, a model with a good `CORR` should help the hedge fund make good returns.\n",
     "\n",
-    "On the Numerai website you will see it referred to as `CORR20V2`, where the `20` denotes that we are using the \"20 day\" version of the target and \"v2\" denotes that we are using the 2nd version of the scoring function. But don't worry too much about all of this for now.  "
+    "On the Numerai website you will see it referred to as `CORR20V2`, where the `20` denotes that we are using the \"20 day\" version of the target and \"v2\" denotes that we are using the 2nd version of the scoring function. But don't worry too much about all of this for now, we will learn more about targets in a future tutorial."
    ]
   },
   {