Add scikit-lego

book/machinelearning/modeltraining.ipynb

@@ -1889,6 +1889,56 @@
     "alpha = [0.05, 0.20]\n",
     "y_pred, y_pis = mapie_regressor.predict(X_test, alpha=alpha)"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "bc7f6fab",
+   "metadata": {},
+   "source": [
+    "## Extra Components For scikit-learn with `scikit-lego`"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "29e041d4",
+   "metadata": {},
+   "source": [
+    "scikit-learn is one of the most popular ML libraries.\n",
+    "\n",
+    "While it's easy to write custom components, it would be nice to have all of them in a single place.\n",
+    "\n",
+    "`scikit-lego` is such a library which contains many custom components like:\n",
+    "\n",
+    "- `DebugPipeline`, which adds debug information to pipelines\n",
+    "- `ImbalancedLinearRegression` to punish over-/underestimation of a model\n",
+    "- `add_lags` to add lag values to a DataFrame\n",
+    "- `ZeroInflatedRegressor` which predicts zero or applies a regression based on a classifier\n",
+    "\n",
+    "and many more!"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "f2c86d37",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.preprocessing import StandardScaler\n",
+    "from sklearn.pipeline import Pipeline\n",
+    "from sklego.preprocessing import RandomAdder\n",
+    "from sklego.mixture import GMMClassifier\n",
+    "\n",
+    "...\n",
+    "\n",
+    "pipeline = Pipeline([\n",
+    "    (\"scale\", StandardScaler()),\n",
+    "    (\"random_noise\", RandomAdder()),\n",
+    "    (\"model\", GMMClassifier())\n",
+    "])\n",
+    "\n",
+    "..."
+   ]
   }
  ],
  "metadata": {