diff --git a/docs/examples/regression_on_workers_compensation.ipynb b/docs/examples/regression_on_workers_compensation.ipynb
new file mode 100644
index 0000000..0d2156c
--- /dev/null
+++ b/docs/examples/regression_on_workers_compensation.ipynb
@@ -0,0 +1,1267 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "0635403e-5936-4c7f-ad25-3a52670a1a10",
+   "metadata": {},
+   "source": [
+    "# Regression on Workers' Compensation Dataset\n",
+    "This notebook demonstrates model diagnostics on the workers' compensation dataset https://www.openml.org/d/42876.\n",
+    "For details, see https://arxiv.org/abs/2202.12780.\n",
+    "\n",
+    "The modelling goal is to predict the expectation $E(Y|X)$ of $Y=\\text{UltimateIncurredClaimCost}$ conditional on the features $X$."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "476f0984-84d4-493e-85c1-9823f2a47491",
+   "metadata": {},
+   "source": [
+    "## 1. Load and prepare data"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "bdf0fa3a-650d-4ea4-a189-e81e4457fd08",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import calendar\n",
+    "import numpy as np\n",
+    "import pandas as pd\n",
+    "from sklearn import set_config\n",
+    "from sklearn.datasets import fetch_openml\n",
+    "\n",
+    "\n",
+    "set_config(transform_output=\"pandas\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "id": "2dcf7ba7-61c1-4ae4-97c7-4968bfc99ee0",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_original = fetch_openml(data_id=42876, parser=\"auto\").frame"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "afe4ccd8-178e-494b-8404-5091ec470fce",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "The prepared dataset contains 82017 rows.\n"
+     ]
+    }
+   ],
+   "source": [
+    "df = df_original.query(\"WeeklyPay >= 200 and HoursWorkedPerWeek >= 20\")\n",
+    "df = df.assign(\n",
+    "    DateTimeOfAccident=lambda x: pd.to_datetime(x[\"DateTimeOfAccident\"]),\n",
+    "    DateOfAccident=lambda x: x.DateTimeOfAccident.dt.date,\n",
+    "    DateReported=lambda x: pd.to_datetime(x.DateReported).dt.date,\n",
+    "    LogDelay=lambda x: np.log1p((x.DateReported - x.DateOfAccident).dt.days),\n",
+    "    HourOfAccident=lambda x: x.DateTimeOfAccident.dt.hour,\n",
+    "    WeekDayOfAccident=lambda x: pd.Categorical.from_codes(\n",
+    "        codes=x.DateTimeOfAccident.dt.weekday,\n",
+    "        dtype=pd.CategoricalDtype(list(calendar.day_name), ordered=True),\n",
+    "    ),\n",
+    "    LogWeeklyPay=lambda x: np.log1p(x.WeeklyPay),\n",
+    "    LogInitial=lambda x: np.log(x.InitialCaseEstimate),\n",
+    "    DependentChildren=lambda x: np.fmin(4, x.DependentChildren),\n",
+    "    HoursWorkedPerWeek=lambda x: np.fmin(60, x.HoursWorkedPerWeek),\n",
+    "    Gender=lambda x: x.Gender.astype(\"category\"),\n",
+    "    MaritalStatus=lambda x: x.MaritalStatus.astype(\"category\"),\n",
+    "    PartTimeFullTime=lambda x: x.PartTimeFullTime.astype(\"category\"),\n",
+    ").rename(columns={\"HoursWorkedPerWeek\": \"HoursPerWeek\"})\n",
+    "\n",
+    "x_continuous = [\n",
+    "    \"Age\",\n",
+    "    \"LogWeeklyPay\",\n",
+    "    \"LogInitial\",\n",
+    "    \"HourOfAccident\",\n",
+    "    \"HoursPerWeek\",\n",
+    "    \"LogDelay\",\n",
+    "]\n",
+    "x_discrete = [\n",
+    "    \"Gender\",\n",
+    "    \"MaritalStatus\",\n",
+    "    \"PartTimeFullTime\",\n",
+    "    \"DependentChildren\",\n",
+    "    \"DaysWorkedPerWeek\",\n",
+    "    \"WeekDayOfAccident\",\n",
+    "]\n",
+    "x_vars = x_continuous + x_discrete\n",
+    "y_var = \"UltimateIncurredClaimCost\"\n",
+    "\n",
+    "print(f\"The prepared dataset contains {df.shape[0]} rows.\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "id": "c23fe013-56a4-4616-b344-8af461593b12",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>DateTimeOfAccident</th>\n",
+       "      <th>DateReported</th>\n",
+       "      <th>Age</th>\n",
+       "      <th>Gender</th>\n",
+       "      <th>MaritalStatus</th>\n",
+       "      <th>DependentChildren</th>\n",
+       "      <th>DependentsOther</th>\n",
+       "      <th>WeeklyPay</th>\n",
+       "      <th>PartTimeFullTime</th>\n",
+       "      <th>HoursPerWeek</th>\n",
+       "      <th>DaysWorkedPerWeek</th>\n",
+       "      <th>ClaimDescription</th>\n",
+       "      <th>InitialCaseEstimate</th>\n",
+       "      <th>UltimateIncurredClaimCost</th>\n",
+       "      <th>DateOfAccident</th>\n",
+       "      <th>LogDelay</th>\n",
+       "      <th>HourOfAccident</th>\n",
+       "      <th>WeekDayOfAccident</th>\n",
+       "      <th>LogWeeklyPay</th>\n",
+       "      <th>LogInitial</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>2005-01-13 09:00:00</td>\n",
+       "      <td>2005-01-24</td>\n",
+       "      <td>45</td>\n",
+       "      <td>M</td>\n",
+       "      <td>S</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>500</td>\n",
+       "      <td>F</td>\n",
+       "      <td>38</td>\n",
+       "      <td>5</td>\n",
+       "      <td>MOVING DISC STRAINED RIGHT SHOULDER</td>\n",
+       "      <td>9500.0</td>\n",
+       "      <td>102.39</td>\n",
+       "      <td>2005-01-13</td>\n",
+       "      <td>2.484907</td>\n",
+       "      <td>9</td>\n",
+       "      <td>Thursday</td>\n",
+       "      <td>6.216606</td>\n",
+       "      <td>9.159047</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>1994-09-28 15:00:00</td>\n",
+       "      <td>1994-10-17</td>\n",
+       "      <td>40</td>\n",
+       "      <td>M</td>\n",
+       "      <td>M</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>283</td>\n",
+       "      <td>F</td>\n",
+       "      <td>38</td>\n",
+       "      <td>5</td>\n",
+       "      <td>BOILING WATER CAME FROM TRUCK STRAIN RIGHT WRIST</td>\n",
+       "      <td>3000.0</td>\n",
+       "      <td>1451.00</td>\n",
+       "      <td>1994-09-28</td>\n",
+       "      <td>2.995732</td>\n",
+       "      <td>15</td>\n",
+       "      <td>Wednesday</td>\n",
+       "      <td>5.648974</td>\n",
+       "      <td>8.006368</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>2002-02-28 07:00:00</td>\n",
+       "      <td>2002-04-08</td>\n",
+       "      <td>50</td>\n",
+       "      <td>F</td>\n",
+       "      <td>S</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>517</td>\n",
+       "      <td>F</td>\n",
+       "      <td>38</td>\n",
+       "      <td>5</td>\n",
+       "      <td>SPILLED CHEMICAL WELDING IRRITATION LEFT CORNEA</td>\n",
+       "      <td>1000.0</td>\n",
+       "      <td>320.28</td>\n",
+       "      <td>2002-02-28</td>\n",
+       "      <td>3.688879</td>\n",
+       "      <td>7</td>\n",
+       "      <td>Thursday</td>\n",
+       "      <td>6.249975</td>\n",
+       "      <td>6.907755</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>1995-04-20 14:00:00</td>\n",
+       "      <td>1995-05-08</td>\n",
+       "      <td>19</td>\n",
+       "      <td>M</td>\n",
+       "      <td>S</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>200</td>\n",
+       "      <td>F</td>\n",
+       "      <td>38</td>\n",
+       "      <td>5</td>\n",
+       "      <td>ENTERED GRINDER FOREIGN BODY RIGHT EYE</td>\n",
+       "      <td>110.0</td>\n",
+       "      <td>108.00</td>\n",
+       "      <td>1995-04-20</td>\n",
+       "      <td>2.944439</td>\n",
+       "      <td>14</td>\n",
+       "      <td>Thursday</td>\n",
+       "      <td>5.303305</td>\n",
+       "      <td>4.700480</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>2005-01-07 14:00:00</td>\n",
+       "      <td>2005-01-31</td>\n",
+       "      <td>19</td>\n",
+       "      <td>M</td>\n",
+       "      <td>S</td>\n",
+       "      <td>0</td>\n",
+       "      <td>0</td>\n",
+       "      <td>767</td>\n",
+       "      <td>F</td>\n",
+       "      <td>40</td>\n",
+       "      <td>5</td>\n",
+       "      <td>LIFTING STRAIN LOWER BACK AND</td>\n",
+       "      <td>9700.0</td>\n",
+       "      <td>7110.90</td>\n",
+       "      <td>2005-01-07</td>\n",
+       "      <td>3.218876</td>\n",
+       "      <td>14</td>\n",
+       "      <td>Friday</td>\n",
+       "      <td>6.643790</td>\n",
+       "      <td>9.179881</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "   DateTimeOfAccident DateReported  Age Gender MaritalStatus  \\\n",
+       "0 2005-01-13 09:00:00   2005-01-24   45      M             S   \n",
+       "1 1994-09-28 15:00:00   1994-10-17   40      M             M   \n",
+       "5 2002-02-28 07:00:00   2002-04-08   50      F             S   \n",
+       "7 1995-04-20 14:00:00   1995-05-08   19      M             S   \n",
+       "8 2005-01-07 14:00:00   2005-01-31   19      M             S   \n",
+       "\n",
+       "   DependentChildren  DependentsOther  WeeklyPay PartTimeFullTime  \\\n",
+       "0                  0                0        500                F   \n",
+       "1                  0                0        283                F   \n",
+       "5                  0                0        517                F   \n",
+       "7                  0                0        200                F   \n",
+       "8                  0                0        767                F   \n",
+       "\n",
+       "   HoursPerWeek  DaysWorkedPerWeek  \\\n",
+       "0            38                  5   \n",
+       "1            38                  5   \n",
+       "5            38                  5   \n",
+       "7            38                  5   \n",
+       "8            40                  5   \n",
+       "\n",
+       "                                   ClaimDescription  InitialCaseEstimate  \\\n",
+       "0               MOVING DISC STRAINED RIGHT SHOULDER               9500.0   \n",
+       "1  BOILING WATER CAME FROM TRUCK STRAIN RIGHT WRIST               3000.0   \n",
+       "5   SPILLED CHEMICAL WELDING IRRITATION LEFT CORNEA               1000.0   \n",
+       "7            ENTERED GRINDER FOREIGN BODY RIGHT EYE                110.0   \n",
+       "8                     LIFTING STRAIN LOWER BACK AND               9700.0   \n",
+       "\n",
+       "   UltimateIncurredClaimCost DateOfAccident  LogDelay  HourOfAccident  \\\n",
+       "0                     102.39     2005-01-13  2.484907               9   \n",
+       "1                    1451.00     1994-09-28  2.995732              15   \n",
+       "5                     320.28     2002-02-28  3.688879               7   \n",
+       "7                     108.00     1995-04-20  2.944439              14   \n",
+       "8                    7110.90     2005-01-07  3.218876              14   \n",
+       "\n",
+       "  WeekDayOfAccident  LogWeeklyPay  LogInitial  \n",
+       "0          Thursday      6.216606    9.159047  \n",
+       "1         Wednesday      5.648974    8.006368  \n",
+       "5          Thursday      6.249975    6.907755  \n",
+       "7          Thursday      5.303305    4.700480  \n",
+       "8            Friday      6.643790    9.179881  "
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df.head()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "1ddf4302-7c1a-4a90-88b6-a00866ce036a",
+   "metadata": {},
+   "source": [
+    "## 2. Data split"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "id": "a541e9f7-e68b-45f6-9b59-96db11c17915",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.model_selection import train_test_split"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "700e6bbb-0641-4141-9d62-b51c78e4427e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_train, df_test = train_test_split(df, train_size=0.75, random_state=1234321)\n",
+    "df = pd.concat((df_train, df_test), axis=0, keys=(\"train\", \"test\")).reset_index(level=0).rename(columns={\"level_0\": \"split\"})\n",
+    "\n",
+    "y_train, y_test = df_train[y_var], df_test[y_var]\n",
+    "X_train, X_test = df_train[x_vars], df_test[x_vars]"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "8c773b7c-f1cd-4deb-ae19-a0a57e1bc30f",
+   "metadata": {},
+   "source": [
+    "We check whether the split results in two about identically distributed samples."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "2c740f0f-a580-4a51-b993-30f78cac2fbc",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>mean</th>\n",
+       "      <th>q20</th>\n",
+       "      <th>q40</th>\n",
+       "      <th>q50</th>\n",
+       "      <th>q60</th>\n",
+       "      <th>q80</th>\n",
+       "      <th>q90</th>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>split</th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "      <th></th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>test</th>\n",
+       "      <td>13300.958856</td>\n",
+       "      <td>205.0</td>\n",
+       "      <td>464.714</td>\n",
+       "      <td>707.500</td>\n",
+       "      <td>1135.984</td>\n",
+       "      <td>4716.426</td>\n",
+       "      <td>18979.432</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>train</th>\n",
+       "      <td>12909.500160</td>\n",
+       "      <td>203.0</td>\n",
+       "      <td>459.000</td>\n",
+       "      <td>691.185</td>\n",
+       "      <td>1096.230</td>\n",
+       "      <td>4596.730</td>\n",
+       "      <td>19030.384</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "               mean    q20      q40      q50       q60       q80        q90\n",
+       "split                                                                      \n",
+       "test   13300.958856  205.0  464.714  707.500  1135.984  4716.426  18979.432\n",
+       "train  12909.500160  203.0  459.000  691.185  1096.230  4596.730  19030.384"
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df.groupby(\"split\").agg(\n",
+    "    mean=pd.NamedAgg(column=y_var, aggfunc=\"mean\"),\n",
+    "    q20=pd.NamedAgg(column=y_var, aggfunc=lambda x: np.quantile(x, 0.2)),\n",
+    "    q40=pd.NamedAgg(column=y_var, aggfunc=lambda x: np.quantile(x, 0.4)),\n",
+    "    q50=pd.NamedAgg(column=y_var, aggfunc=lambda x: np.quantile(x, 0.5)),\n",
+    "    q60=pd.NamedAgg(column=y_var, aggfunc=lambda x: np.quantile(x, 0.6)),\n",
+    "    q80=pd.NamedAgg(column=y_var, aggfunc=lambda x: np.quantile(x, 0.8)),\n",
+    "    q90=pd.NamedAgg(column=y_var, aggfunc=lambda x: np.quantile(x, 0.9)),\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "1e32a519-7ac8-4e94-99a6-f451dc2190ff",
+   "metadata": {},
+   "source": [
+    "## 3. Models\n",
+    "We aim at predicting the conditional expectation $E(Y|X)$ and will finally evaluate the models with the Gamma deviance $S(z, y) = 2 \\left(log(\\frac{z}{y}) + \\frac{y}{z} - 1\\right)$.\n",
+    "### 3.1 The trivial model"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "id": "ad0611d6-3f25-4fdd-9ae4-f6f3161233cb",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.dummy import DummyRegressor\n",
+    "\n",
+    "\n",
+    "m_trivial = DummyRegressor(strategy=\"mean\").fit(X_train, y_train)"
+   ]
+  },
+  {
+   "attachments": {},
+   "cell_type": "markdown",
+   "id": "9a176508-87d9-4b45-b9c1-436becc205f3",
+   "metadata": {},
+   "source": [
+    "## 3.2 OLS\n",
+    "\n",
+    "Here, we train a Ordinary Least Squares (OLS) model, but on the log transformed target.\n",
+    "`TransformedTargetRegressor` takes care of this transformation and also of the back transformation such that predictions are on the original target UltimateIncurredClaimCost."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "id": "70f55852-de16-4dc6-b35e-ad3dfef93b3e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.compose import ColumnTransformer, TransformedTargetRegressor\n",
+    "from sklearn.linear_model import LinearRegression\n",
+    "from sklearn.pipeline import Pipeline\n",
+    "from sklearn.preprocessing import OneHotEncoder\n",
+    "\n",
+    "\n",
+    "# ColumnTransformer for linear models\n",
+    "col_trans_linear = ColumnTransformer(\n",
+    "    [\n",
+    "        (\"numeric_features\", \"passthrough\", x_continuous),\n",
+    "        (\"categorical_features\", OneHotEncoder(drop=\"first\", sparse_output=False), x_discrete),\n",
+    "    ]\n",
+    ")\n",
+    "\n",
+    "m_ols = Pipeline(\n",
+    "    [\n",
+    "        (\"column_transformer\", col_trans_linear),\n",
+    "        (\n",
+    "            \"model\",\n",
+    "            TransformedTargetRegressor(\n",
+    "                regressor=LinearRegression(),\n",
+    "                func=np.log,\n",
+    "                inverse_func=np.exp,\n",
+    "            )\n",
+    "        ),\n",
+    "    ]\n",
+    ").fit(X_train, y_train)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "39651cc8-ae00-4699-9c77-3846e1658c16",
+   "metadata": {},
+   "source": [
+    "We expect a high bias due to the transformation. We correct for this bias by a multiplicative constant such that on the training set: $\\sum_i m(x_i) = \\sum_i y_i$."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "id": "b76d3139-c4bd-44ad-ac64-42b8323f51fc",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "The correction factor is 6.6446580674943165\n"
+     ]
+    }
+   ],
+   "source": [
+    "ols_corr_factor = np.sum(y_train) / np.sum(m_ols.predict(X_train))\n",
+    "print(f\"The correction factor is {ols_corr_factor}\")\n",
+    "m_ols[-1].regressor_.intercept_ += np.log(ols_corr_factor)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "id": "658210ee-ddca-4017-9572-a0c5797321d1",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "1.0000000000000002"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "np.sum(y_train) / np.sum(m_ols.predict(X_train))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "27244ec1-67ca-47f4-9820-2caade7386ce",
+   "metadata": {},
+   "source": [
+    "### 3.3 Poisson GLM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "id": "66149c6d-8760-4769-8201-020df64ee75b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.linear_model import PoissonRegressor\n",
+    "\n",
+    "\n",
+    "m_glm_poisson = Pipeline(\n",
+    "    [\n",
+    "        (\"column_transformer\", col_trans_linear),\n",
+    "        (\"model\", PoissonRegressor(max_iter=10_000, alpha=1e-15, solver=\"newton-cholesky\")),\n",
+    "    ]\n",
+    ").fit(X_train, y_train)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "82daee1e-b891-4589-8d62-7895d0aa2d8c",
+   "metadata": {},
+   "source": [
+    "### 3.4 Gradient Boosted Decision Trees\n",
+    "TODO: We would like to use Gamma deviance as loss, but scikit-learn does not (yet) support it. We could use XGBoost or LightGBM instead, but are content with the Poisson loss at the moment."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "id": "c791a0b8-2218-4c4f-98dc-7e048b3ff76e",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "CPU times: user 13min 36s, sys: 27.6 s, total: 14min 3s\n",
+      "Wall time: 2min 2s\n"
+     ]
+    }
+   ],
+   "source": [
+    "%%time\n",
+    "# Note that this cell might take a little while ~ 2 minutes on my laptop.\n",
+    "from sklearn.ensemble import HistGradientBoostingRegressor\n",
+    "from sklearn.experimental import enable_halving_search_cv\n",
+    "from sklearn.model_selection import HalvingGridSearchCV\n",
+    "from sklearn.preprocessing import OrdinalEncoder\n",
+    "\n",
+    "\n",
+    "# ColumnTransformer for boosted trees\n",
+    "col_trans_bt = ColumnTransformer(\n",
+    "    [\n",
+    "        (\"categorical_features\", OrdinalEncoder(), x_discrete),\n",
+    "        (\"numeric_features\", \"passthrough\", x_continuous),\n",
+    "    ]\n",
+    ")\n",
+    "m_hgbt_poisson_base = Pipeline(\n",
+    "    [\n",
+    "        (\"column_transformer\", col_trans_bt),\n",
+    "        (\n",
+    "            \"model\", HistGradientBoostingRegressor(\n",
+    "                loss=\"poisson\",\n",
+    "                categorical_features=list(range(len(x_discrete))),\n",
+    "                monotonic_cst={\"numeric_features__LogWeeklyPay\": 1},  # set_output API, YES!\n",
+    "                max_iter=200,\n",
+    "                random_state=33,\n",
+    "            )\n",
+    "        ),\n",
+    "    ]\n",
+    ")\n",
+    "\n",
+    "param_grid = {\n",
+    "    \"model__learning_rate\": [0.02, 0.05],\n",
+    "    \"model__min_samples_leaf\": [20, 30, 40],\n",
+    "    \"model__l2_regularization\": [0.1, 1],\n",
+    "    \"model__max_depth\": [None, 3],\n",
+    "}\n",
+    "\n",
+    "\n",
+    "# successive halfing grid search (CV) on the training data\n",
+    "shgs = HalvingGridSearchCV(\n",
+    "    m_hgbt_poisson_base,\n",
+    "    param_grid=param_grid,\n",
+    "    cv=5,\n",
+    "    scoring=\"neg_mean_gamma_deviance\",\n",
+    "    random_state=321,\n",
+    ").fit(X_train, y_train)\n",
+    "m_hgbt_poisson = shgs.best_estimator_"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "id": "3ee2c06c-8e85-40f6-9133-727e2af81cab",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "({'model__l2_regularization': 1,\n",
+       "  'model__learning_rate': 0.05,\n",
+       "  'model__max_depth': 3,\n",
+       "  'model__min_samples_leaf': 40},\n",
+       " -3.547907637012485)"
+      ]
+     },
+     "execution_count": 14,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "shgs.best_params_, shgs.best_score_"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "id": "6dacffc8-9fd8-4b56-bd4e-9a9b71dbbdcb",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "130"
+      ]
+     },
+     "execution_count": 15,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "m_hgbt_poisson[-1].n_iter_"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "1a5f12d6-966d-45cc-ac1e-cfd43595a081",
+   "metadata": {},
+   "source": [
+    "## 4. Calibration Assessment\n",
+    "To make the code easier, we put together the predictions of our models on the test set in a single dataframe."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "id": "4bd40f96-2581-499e-8bc6-2a3bb019de7b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_pred_test = pd.DataFrame(\n",
+    "    {\n",
+    "        \"Trivial\": m_trivial.predict(X_test),\n",
+    "        \"OLS\": m_ols.predict(X_test),\n",
+    "        \"GLM_Poisson\": m_glm_poisson.predict(X_test),\n",
+    "        \"HGBT_Poisson\": m_hgbt_poisson.predict(X_test),\n",
+    "    }\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "631c120d-c991-4c51-9a49-0282c21afbf4",
+   "metadata": {},
+   "source": [
+    "## 4.1. Reliability Diagrams\n",
+    "A reliability diagram plots an estimation of $E(Y|m(X))$ versus $m(X)$.\n",
+    "A good way to estimate $E(Y|m(X))$ and thereby avoiding manual binning is by isotonic regression.\n",
+    "This is implemented in `plot_reliability_diagram`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "id": "70fa98bf-2ab2-40fe-a1ca-ce3af33ea261",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import matplotlib.pyplot as plt\n",
+    "from model_diagnostics.calibration import compute_bias, plot_bias, plot_reliability_diagram"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "id": "f58490f0-9cae-4f2d-bf89-7f7a501d9995",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Text(0.5, 1.0, 'Reliability Diagram HGBT_Poisson Poisson HGBT')"
+      ]
+     },
+     "execution_count": 18,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 1200x400 with 3 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "fig, axes = plt.subplots(ncols=3, figsize=(12, 4), sharex=True, sharey=True)\n",
+    "\n",
+    "plot_reliability_diagram(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=df_pred_test[\"OLS\"],\n",
+    "    ax=axes[0],\n",
+    ")\n",
+    "axes[0].set_title(axes[0].get_title() + \" OLS\")\n",
+    "\n",
+    "plot_reliability_diagram(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=df_pred_test[\"GLM_Poisson\"],\n",
+    "    ax=axes[1]\n",
+    ")\n",
+    "axes[1].set_title(axes[1].get_title() + \" Poisson GLM\")\n",
+    "\n",
+    "plot_reliability_diagram(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=df_pred_test[\"HGBT_Poisson\"],\n",
+    "    ax=axes[2]\n",
+    ")\n",
+    "axes[2].set_title(axes[2].get_title() + \" Poisson HGBT\")\n",
+    "#fig.set_layout_engine(layout=\"tight\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6e1789fb-b36c-4e37-b764-29ca5e031bee",
+   "metadata": {},
+   "source": [
+    "### 4.2 Conditional Bias"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "a77ba53f-dea6-4b54-afd2-2d5fcd39ac77",
+   "metadata": {},
+   "source": [
+    "### Unconditional Bias"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "id": "966a21af-8b3e-4d60-a25e-4e1334afeadc",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>model</th>\n",
+       "      <th>bias_mean</th>\n",
+       "      <th>bias_count</th>\n",
+       "      <th>bias_stddev</th>\n",
+       "      <th>p_value</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>Trivial</td>\n",
+       "      <td>-391.458696</td>\n",
+       "      <td>20505</td>\n",
+       "      <td>56480.368800</td>\n",
+       "      <td>0.320979</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>OLS</td>\n",
+       "      <td>-158.226277</td>\n",
+       "      <td>20505</td>\n",
+       "      <td>53049.749161</td>\n",
+       "      <td>0.669314</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>GLM_Poisson</td>\n",
+       "      <td>-142.672362</td>\n",
+       "      <td>20505</td>\n",
+       "      <td>53048.066624</td>\n",
+       "      <td>0.700150</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>HGBT_Poisson</td>\n",
+       "      <td>-94.439905</td>\n",
+       "      <td>20505</td>\n",
+       "      <td>53100.300161</td>\n",
+       "      <td>0.798976</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "          model   bias_mean  bias_count   bias_stddev   p_value\n",
+       "0       Trivial -391.458696       20505  56480.368800  0.320979\n",
+       "1           OLS -158.226277       20505  53049.749161  0.669314\n",
+       "2   GLM_Poisson -142.672362       20505  53048.066624  0.700150\n",
+       "3  HGBT_Poisson  -94.439905       20505  53100.300161  0.798976"
+      ]
+     },
+     "execution_count": 19,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "compute_bias(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=df_pred_test,\n",
+    "    feature=None,\n",
+    ").to_pandas()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c828b8d2-1b80-4561-8d1d-0e24e5f6f85c",
+   "metadata": {},
+   "source": [
+    "#### Bias Conditional on Gender"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "id": "bcc0aaf6-8653-4a7d-8c66-c0a4713e700f",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>model</th>\n",
+       "      <th>Gender</th>\n",
+       "      <th>bias_mean</th>\n",
+       "      <th>bias_count</th>\n",
+       "      <th>bias_stddev</th>\n",
+       "      <th>p_value</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>Trivial</td>\n",
+       "      <td>F</td>\n",
+       "      <td>-4693.414762</td>\n",
+       "      <td>4441</td>\n",
+       "      <td>61676.821475</td>\n",
+       "      <td>4.115092e-07</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>1</th>\n",
+       "      <td>Trivial</td>\n",
+       "      <td>M</td>\n",
+       "      <td>797.845767</td>\n",
+       "      <td>16064</td>\n",
+       "      <td>54899.756026</td>\n",
+       "      <td>6.550241e-02</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>OLS</td>\n",
+       "      <td>F</td>\n",
+       "      <td>-1212.773117</td>\n",
+       "      <td>4441</td>\n",
+       "      <td>58523.230496</td>\n",
+       "      <td>1.673504e-01</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>OLS</td>\n",
+       "      <td>M</td>\n",
+       "      <td>133.310235</td>\n",
+       "      <td>16064</td>\n",
+       "      <td>51432.053128</td>\n",
+       "      <td>7.425259e-01</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>GLM_Poisson</td>\n",
+       "      <td>F</td>\n",
+       "      <td>-184.995122</td>\n",
+       "      <td>4441</td>\n",
+       "      <td>58572.701607</td>\n",
+       "      <td>8.333048e-01</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>GLM_Poisson</td>\n",
+       "      <td>M</td>\n",
+       "      <td>-130.971952</td>\n",
+       "      <td>16064</td>\n",
+       "      <td>51418.078829</td>\n",
+       "      <td>7.468195e-01</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>HGBT_Poisson</td>\n",
+       "      <td>F</td>\n",
+       "      <td>-303.384575</td>\n",
+       "      <td>4441</td>\n",
+       "      <td>58351.353739</td>\n",
+       "      <td>7.289956e-01</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>HGBT_Poisson</td>\n",
+       "      <td>M</td>\n",
+       "      <td>-36.675756</td>\n",
+       "      <td>16064</td>\n",
+       "      <td>51556.136879</td>\n",
+       "      <td>9.281593e-01</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "          model Gender    bias_mean  bias_count   bias_stddev       p_value\n",
+       "0       Trivial      F -4693.414762        4441  61676.821475  4.115092e-07\n",
+       "1       Trivial      M   797.845767       16064  54899.756026  6.550241e-02\n",
+       "2           OLS      F -1212.773117        4441  58523.230496  1.673504e-01\n",
+       "3           OLS      M   133.310235       16064  51432.053128  7.425259e-01\n",
+       "4   GLM_Poisson      F  -184.995122        4441  58572.701607  8.333048e-01\n",
+       "5   GLM_Poisson      M  -130.971952       16064  51418.078829  7.468195e-01\n",
+       "6  HGBT_Poisson      F  -303.384575        4441  58351.353739  7.289956e-01\n",
+       "7  HGBT_Poisson      M   -36.675756       16064  51556.136879  9.281593e-01"
+      ]
+     },
+     "execution_count": 20,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "compute_bias(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=df_pred_test,\n",
+    "    feature=X_test[\"Gender\"],\n",
+    ").to_pandas()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "id": "a57d6452-5f93-409b-82e2-36350fa86545",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<AxesSubplot: title={'center': 'Bias Plot'}, xlabel='Gender', ylabel='bias'>"
+      ]
+     },
+     "execution_count": 21,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plot_bias(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=m_ols.predict(X_test),\n",
+    "    feature=X_test[\"Gender\"],\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "1ac67348-f61a-49fe-b807-66b640e0c298",
+   "metadata": {},
+   "source": [
+    "#### Bias Conditional on LogWeeklyPay"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "id": "8c1eade0-cbc1-4919-8edd-a380491cc03d",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<AxesSubplot: title={'center': 'Bias Plot'}, xlabel='binned LogWeeklyPay', ylabel='bias'>"
+      ]
+     },
+     "execution_count": 22,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plot_bias(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=m_ols.predict(X_test),\n",
+    "    feature=X_test[\"LogWeeklyPay\"],\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "id": "b17d2803-158c-4467-87f0-c008b66eb18b",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<AxesSubplot: title={'center': 'Bias Plot'}, xlabel='binned LogWeeklyPay', ylabel='bias'>"
+      ]
+     },
+     "execution_count": 23,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plot_bias(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=m_hgbt_poisson.predict(X_test),\n",
+    "    feature=X_test[\"LogWeeklyPay\"],\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "e969d6df-c923-4733-b256-312e08a5ebf3",
+   "metadata": {},
+   "source": [
+    "#### Bias Conditional on Age"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "id": "ed065707-1d71-41e0-8b4e-f70b4478c7db",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<AxesSubplot: title={'center': 'Bias Plot'}, xlabel='binned Age', ylabel='bias'>"
+      ]
+     },
+     "execution_count": 24,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 640x480 with 1 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "plot_bias(\n",
+    "    y_obs=y_test,\n",
+    "    y_pred=m_ols.predict(X_test),\n",
+    "    feature=X_test[\"Age\"],\n",
+    ")"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "jupyter",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.10.8"
+  },
+  "vscode": {
+   "interpreter": {
+    "hash": "96b0fa0ec69a2f08511d770c35d8e462f16e1f34df81292631ff071c65461c9e"
+   }
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/mkdocs.yml b/mkdocs.yml
index fceb535..4f688ae 100644
--- a/mkdocs.yml
+++ b/mkdocs.yml
@@ -32,6 +32,8 @@ theme:
 nav:
   - Home:
     - About: index.md
+  - Examples:
+    - Regression on Workers' Compensation: examples/regression_on_workers_compensation.ipynb
   - Code Reference: reference/  # defer to gen-files + literate-nav
 
 watch: