diff --git a/models/model.joblib b/models/model.joblib
index 9b39aee..533a414 100644
Binary files a/models/model.joblib and b/models/model.joblib differ
diff --git a/notebooks/02-comparative_analysis.ipynb b/notebooks/02-comparative_analysis.ipynb
index 68c85a0..2dd0a67 100644
--- a/notebooks/02-comparative_analysis.ipynb
+++ b/notebooks/02-comparative_analysis.ipynb
@@ -8,6 +8,34 @@
     "# Análise comparativa"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "713e7038-de8e-45b1-ae90-f748f7c823ee",
+   "metadata": {},
+   "source": [
+    "## Objetivo:\n",
+    "O objetivo deste notebook é realizar uma análise comparativa de diferentes modelos de aprendizado de máquina aplicados à previsão da demanda por compartilhamento/locação de bicicletas. A análise busca identificar quais modelos apresentam o melhor desempenho em termos de previsão, levando em consideração a importância da preparação adequada dos dados e a aplicação de técnicas de validação robustas."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "a344d8d1-8d0b-458e-bac1-2438cc8a0186",
+   "metadata": {},
+   "source": [
+    "## Contexto:\n",
+    "A locação de bicicletas é uma solução de mobilidade urbana sustentável que tem ganhado popularidade em muitas cidades. No entanto, a previsão precisa da demanda por bicicletas é crucial para otimizar a disponibilidade e reduzir custos operacionais. Este notebook aborda o desafio de prever a demanda utilizando uma base de dados real de locação de bicicletas. A preparação dos dados e a escolha de modelos adequados são etapas fundamentais para garantir a precisão das previsões. A comparação entre diferentes modelos permitirá identificar a abordagem mais eficaz para este problema. O conjunto de dados \"[Bike Sharing](https://www.kaggle.com/datasets/lakshmi25npathi/bike-sharing-dataset/data)\" do Kaggle registra o uso de bicicletas compartilhadas em Washington, D.C. entre o início de 2011 e 2013. Ele inclui variáveis como data, estação do ano, se o dia é feriado ou útil, situação climática, temperatura, sensação térmica, umidade, velocidade do vento, além do número de usuários casuais, registrados e o total de aluguéis. Essas informações permitem analisar padrões de uso e construir modelos para prever a demanda de bicicletas com base em condições climáticas e temporais.\n",
+    "\n",
+    "---"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "f7f6bdb1-195d-42a4-a109-6156e33b23c2",
+   "metadata": {},
+   "source": [
+    "## Configuração Inicial:"
+   ]
+  },
   {
    "cell_type": "code",
    "execution_count": 1,
@@ -21,13 +49,15 @@
     "import joblib\n",
     "import numpy as np\n",
     "import pandas as pd\n",
+    "import seaborn as sns\n",
+    "import matplotlib.pyplot as plt\n",
     "\n",
     "from sklearn.pipeline import Pipeline\n",
     "from sklearn.impute import SimpleImputer\n",
     "from sklearn.preprocessing import StandardScaler, OneHotEncoder, OrdinalEncoder\n",
     "from sklearn.compose import ColumnTransformer\n",
     "\n",
-    "from sklearn.model_selection import ShuffleSplit, GridSearchCV, KFold, cross_validate\n",
+    "from sklearn.model_selection import ShuffleSplit, GridSearchCV, KFold, cross_validate, train_test_split, learning_curve\n",
     "from sklearn.linear_model import LinearRegression\n",
     "from sklearn.neighbors import KNeighborsRegressor\n",
     "from sklearn.ensemble import GradientBoostingRegressor, RandomForestRegressor\n",
@@ -39,7 +69,23 @@
    "id": "d6821558-3c26-4b1b-b2ac-8ef422219bb8",
    "metadata": {},
    "source": [
-    "## 1. Obtenção dos dados:"
+    "## 1. Preparação dos Dados:"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "8cd15780-f875-4cb2-ab70-07461299672f",
+   "metadata": {},
+   "source": [
+    "A preparação dos dados é uma etapa essencial no desenvolvimento de modelos de aprendizado de máquina, pois garante que os dados estejam em um formato adequado para a análise e modelagem. Nesta etapa, serão realizados os seguintes processos: Tratamento de dados faltantes e discrepantes, Codificação de variáveis e Normalização dos dados."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "af991dff-12ed-4ccb-9ddb-a3c85bd8bd8b",
+   "metadata": {},
+   "source": [
+    "### 1.1 Obtenção de dados"
    ]
   },
   {
@@ -246,10 +292,22 @@
   },
   {
    "cell_type": "markdown",
-   "id": "3e67c7e6-c133-4ff8-871a-d4e1a8e132b7",
+   "id": "30be6d70-b36e-4b55-aba6-5dcfc60bd2d2",
    "metadata": {},
    "source": [
-    "---"
+    "- A coluna `instant` trata somente do índice, então podemos removê-la."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "id": "78247379-94c0-4319-abb2-c2629f28d803",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [],
+   "source": [
+    "df = df.drop(columns=['instant'])"
    ]
   },
   {
@@ -257,12 +315,12 @@
    "id": "2e0c1fc4-8764-4347-8389-8ad0b7d3291d",
    "metadata": {},
    "source": [
-    "## 2. Preparação de dados:"
+    "### 1.2 Tratamento de dados Faltantes"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 4,
    "id": "3ee02410-060d-482d-ae22-1efce480dc8e",
    "metadata": {
     "tags": []
@@ -273,7 +331,6 @@
      "output_type": "stream",
      "text": [
       "Dados faltantes por coluna:\n",
-      "instant       0\n",
       "dteday        0\n",
       "season        0\n",
       "yr            0\n",
@@ -298,9 +355,299 @@
     "print(df.isnull().sum())"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "id": "2b481836-aa87-4c63-9d0f-5d5e925e4ebc",
+   "metadata": {
+    "tags": []
+   },
+   "source": [
+    "Não há dados faltantes para nenhuma das colunas na base de dados."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "a847829b-43fd-412f-bcc0-402cf64738e6",
+   "metadata": {},
+   "source": [
+    "### 1.3 Identificação de Outliers"
+   ]
+  },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 5,
+   "id": "941da82c-d120-458e-8042-47297e1e542a",
+   "metadata": {
+    "tags": []
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "'temp': 0 outliers\n",
+      "'atemp': 0 outliers\n",
+      "'hum': 2 outliers\n",
+      "'windspeed': 13 outliers\n",
+      "'casual': 44 outliers\n",
+      "'registered': 0 outliers\n",
+      "'cnt': 0 outliers\n"
+     ]
+    }
+   ],
+   "source": [
+    "def detect_outliers_iqr(df):\n",
+    "    cols_to_check = ['temp', 'atemp', 'hum', 'windspeed', 'casual', 'registered', 'cnt']\n",
+    "    outliers_dict = {}\n",
+    "\n",
+    "    for col in cols_to_check:\n",
+    "        Q1 = df[col].quantile(0.25)\n",
+    "        Q3 = df[col].quantile(0.75)\n",
+    "        IQR = Q3 - Q1\n",
+    "        lower_bound = Q1 - 1.5 * IQR\n",
+    "        upper_bound = Q3 + 1.5 * IQR\n",
+    "\n",
+    "        num_outliers = ((df[col] < lower_bound) | (df[col] > upper_bound)).sum()\n",
+    "        outliers_dict[col] = num_outliers\n",
+    "\n",
+    "    for col, num_outliers in outliers_dict.items():\n",
+    "        print(f\"'{col}': {num_outliers} outliers\")\n",
+    "\n",
+    "detect_outliers_iqr(df)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "3f1fbc3f-c841-45f6-8eb6-653ecbe09298",
+   "metadata": {},
+   "source": [
+    "### 1.4 Codificação de Variáveis"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "id": "49eb2f04-9750-4759-a079-daa9fec511ff",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df['dteday'] = pd.to_datetime(df['dteday'])\n",
+    "\n",
+    "df['day'] = df['dteday'].dt.day\n",
+    "\n",
+    "df = df.drop(columns=['dteday'])"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "d3161acf-8a8f-473e-b5c0-04467c5e0c63",
+   "metadata": {},
+   "source": [
+    "Como a base de dados já possui variáveis referentes a mês (`mnth`) e ano(`yr`), separamos apenas o dia. Com isso, precisamos alterar nosso dicionário de dados."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "id": "4ad83997-721d-4d24-93ad-bb30f035c126",
+   "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>variavel</th>\n",
+       "      <th>descricao</th>\n",
+       "      <th>tipo</th>\n",
+       "      <th>subtipo</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>2</th>\n",
+       "      <td>season</td>\n",
+       "      <td>Estação do ano</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>ordinal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>3</th>\n",
+       "      <td>yr</td>\n",
+       "      <td>Ano</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>ordinal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>4</th>\n",
+       "      <td>mnth</td>\n",
+       "      <td>Mês</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>ordinal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>5</th>\n",
+       "      <td>holiday</td>\n",
+       "      <td>Se o dia é feriado ou não</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>nominal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>6</th>\n",
+       "      <td>weekday</td>\n",
+       "      <td>Dia da semana</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>nominal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>7</th>\n",
+       "      <td>workingday</td>\n",
+       "      <td>Se o dia não é fim de semana e nem feriado</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>nominal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>8</th>\n",
+       "      <td>weathersit</td>\n",
+       "      <td>Clima</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>nominal</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>9</th>\n",
+       "      <td>temp</td>\n",
+       "      <td>Temperatura normalizada em Celsius</td>\n",
+       "      <td>quantitativa</td>\n",
+       "      <td>contínua</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>10</th>\n",
+       "      <td>atemp</td>\n",
+       "      <td>Temperatura de sensação normalizada em Celsius</td>\n",
+       "      <td>quantitativa</td>\n",
+       "      <td>contínua</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>11</th>\n",
+       "      <td>hum</td>\n",
+       "      <td>Umidade do ar</td>\n",
+       "      <td>quantitiva</td>\n",
+       "      <td>contínua</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12</th>\n",
+       "      <td>windspeed</td>\n",
+       "      <td>Velocidade do vento</td>\n",
+       "      <td>quantitiva</td>\n",
+       "      <td>contínua</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13</th>\n",
+       "      <td>casual</td>\n",
+       "      <td>Contagem de usuários casuais</td>\n",
+       "      <td>quantitiva</td>\n",
+       "      <td>discreta</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>14</th>\n",
+       "      <td>registered</td>\n",
+       "      <td>Contagem de usuários registrados</td>\n",
+       "      <td>quantitiva</td>\n",
+       "      <td>discreta</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>15</th>\n",
+       "      <td>cnt</td>\n",
+       "      <td>Contagem total de bicicletas alugadas, incluin...</td>\n",
+       "      <td>quantitiva</td>\n",
+       "      <td>discreta</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>16</th>\n",
+       "      <td>day</td>\n",
+       "      <td>Dia no ano</td>\n",
+       "      <td>qualitativa</td>\n",
+       "      <td>ordinal</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "      variavel                                          descricao  \\\n",
+       "2       season                                     Estação do ano   \n",
+       "3           yr                                                Ano   \n",
+       "4         mnth                                                Mês   \n",
+       "5      holiday                          Se o dia é feriado ou não   \n",
+       "6      weekday                                      Dia da semana   \n",
+       "7   workingday         Se o dia não é fim de semana e nem feriado   \n",
+       "8   weathersit                                              Clima   \n",
+       "9         temp                 Temperatura normalizada em Celsius   \n",
+       "10       atemp     Temperatura de sensação normalizada em Celsius   \n",
+       "11         hum                                      Umidade do ar   \n",
+       "12   windspeed                                Velocidade do vento   \n",
+       "13      casual                       Contagem de usuários casuais   \n",
+       "14  registered                   Contagem de usuários registrados   \n",
+       "15         cnt  Contagem total de bicicletas alugadas, incluin...   \n",
+       "16         day                                         Dia no ano   \n",
+       "\n",
+       "            tipo   subtipo  \n",
+       "2    qualitativa   ordinal  \n",
+       "3    qualitativa   ordinal  \n",
+       "4    qualitativa   ordinal  \n",
+       "5    qualitativa   nominal  \n",
+       "6    qualitativa   nominal  \n",
+       "7    qualitativa   nominal  \n",
+       "8    qualitativa   nominal  \n",
+       "9   quantitativa  contínua  \n",
+       "10  quantitativa  contínua  \n",
+       "11    quantitiva  contínua  \n",
+       "12    quantitiva  contínua  \n",
+       "13    quantitiva  discreta  \n",
+       "14    quantitiva  discreta  \n",
+       "15    quantitiva  discreta  \n",
+       "16   qualitativa   ordinal  "
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "new_columns = pd.DataFrame({\n",
+    "    'variavel': ['day'],\n",
+    "    'descricao': [\n",
+    "        'Dia no ano'\n",
+    "    ],\n",
+    "    'tipo': ['qualitativa'],\n",
+    "    'subtipo': ['ordinal']\n",
+    "})\n",
+    "\n",
+    "df_dict = pd.concat([df_dict, new_columns], ignore_index=True)\n",
+    "df_dict = df_dict[~df_dict['variavel'].isin(['instant', 'dteday'])]\n",
+    "df_dict"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
    "id": "978ca630-47b5-4011-b108-ab5b6492ec6e",
    "metadata": {
     "tags": []
@@ -325,7 +672,7 @@
     "\n",
     "ordinal_columns = (\n",
     "    df_dict\n",
-    "    .query(\"subtipo == 'ordinal' and variavel != 'dteday'\")\n",
+    "    .query(\"subtipo == 'ordinal'\")\n",
     "    .variavel\n",
     "    .to_list()\n",
     ")\n",
@@ -343,23 +690,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 9,
    "id": "3f0346d7-3228-44af-bfb5-12f0712979ee",
    "metadata": {
     "tags": []
    },
    "outputs": [],
    "source": [
-    "if 'dteday' in X.columns:\n",
-    "    X['year'] = X['dteday'].dt.year\n",
-    "    X['month'] = X['dteday'].dt.month\n",
-    "    X['day'] = X['dteday'].dt.day\n",
-    "    X = X.drop(columns=['dteday'])\n",
-    "\n",
     "nominal_preprocessor = Pipeline([\n",
     "    ('missing', SimpleImputer(strategy='most_frequent')), \n",
-    "    ('encoding', OneHotEncoder(sparse_output=False, drop='first')), \n",
-    "    ('normalization', StandardScaler())  \n",
+    "    ('encoding', OneHotEncoder(sparse_output=False, drop='first')),\n",
+    "    ('normalization', StandardScaler())\n",
     "])\n",
     "\n",
     "continuous_preprocessor = Pipeline([\n",
@@ -396,31 +737,48 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 10,
    "id": "e144b0d5-2285-49c9-95b2-15a796e179da",
    "metadata": {
     "tags": []
    },
    "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Colunas após a preparação:\n",
+      "['nominal__holiday_1' 'nominal__weekday_1' 'nominal__weekday_2'\n",
+      " 'nominal__weekday_3' 'nominal__weekday_4' 'nominal__weekday_5'\n",
+      " 'nominal__weekday_6' 'nominal__workingday_1' 'nominal__weathersit_2'\n",
+      " 'nominal__weathersit_3' 'ordinal__season' 'ordinal__yr' 'ordinal__mnth'\n",
+      " 'ordinal__day' 'discrete__casual' 'discrete__registered'\n",
+      " 'continuous__temp' 'continuous__atemp' 'continuous__hum'\n",
+      " 'continuous__windspeed']\n"
+     ]
+    },
     {
      "data": {
       "text/plain": [
        "(731, 20)"
       ]
      },
-     "execution_count": 6,
+     "execution_count": 10,
      "metadata": {},
      "output_type": "execute_result"
     }
    ],
    "source": [
     "X_prepared = preprocessor.fit_transform(X)\n",
+    "feature_names = preprocessor.get_feature_names_out()\n",
+    "print(\"Colunas após a preparação:\")\n",
+    "print(feature_names)\n",
     "X_prepared.shape"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 11,
    "id": "e4a589c6-aebc-42b7-8b96-b564784db2b8",
    "metadata": {
     "tags": []
@@ -466,7 +824,7 @@
    "id": "33e33101-a682-4cc2-9adf-362464aa7791",
    "metadata": {},
    "source": [
-    "## 3. Seleção de modelos\n"
+    "## 2. Seleção de modelos"
    ]
   },
   {
@@ -492,7 +850,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 12,
    "id": "d18ec4ad-8cd7-408e-b700-7bb953934592",
    "metadata": {
     "tags": []
@@ -509,8 +867,9 @@
     "max_iter = 1000 \n",
     "models = [\n",
     "    ('K-Nearest Neighbors', KNeighborsRegressor(), {\n",
-    "        \"n_neighbors\": range(3, 20, 2), \n",
-    "        'weights': ['uniform', 'distance']\n",
+    "        \"n_neighbors\": range(3, 5, 7), \n",
+    "        'weights': ['uniform', 'distance'],\n",
+    "        'algorithm': ['auto', 'ball_tree', 'kd_tree', 'brute']\n",
     "    }),\n",
     "    ('Gradient Boosting', GradientBoostingRegressor(random_state=random_state), {\n",
     "        'n_estimators': [50, 100, 150],\n",
@@ -531,7 +890,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 13,
    "id": "38a1312c-22be-4db7-89a4-12a69fe2c940",
    "metadata": {
     "tags": []
@@ -570,27 +929,24 @@
        "      <th>test_neg_mean_squared_error</th>\n",
        "      <th>test_neg_mean_absolute_error</th>\n",
        "      <th>test_r2</th>\n",
-       "      <th>model_name</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>mean</th>\n",
-       "      <td>0.778052</td>\n",
-       "      <td>0.029866</td>\n",
-       "      <td>-495085.285587</td>\n",
-       "      <td>-521.421224</td>\n",
-       "      <td>0.867283</td>\n",
-       "      <td>K-Nearest Neighbors</td>\n",
+       "      <td>0.958916</td>\n",
+       "      <td>0.086565</td>\n",
+       "      <td>-520945.104332</td>\n",
+       "      <td>-538.463882</td>\n",
+       "      <td>0.859563</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>std</th>\n",
-       "      <td>0.125487</td>\n",
-       "      <td>0.014388</td>\n",
-       "      <td>98807.812136</td>\n",
-       "      <td>56.917455</td>\n",
-       "      <td>0.024541</td>\n",
-       "      <td>K-Nearest Neighbors</td>\n",
+       "      <td>0.314371</td>\n",
+       "      <td>0.049576</td>\n",
+       "      <td>80345.848916</td>\n",
+       "      <td>36.727944</td>\n",
+       "      <td>0.024787</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -598,12 +954,12 @@
       ],
       "text/plain": [
        "      fit_time  score_time  test_neg_mean_squared_error  \\\n",
-       "mean  0.778052    0.029866               -495085.285587   \n",
-       "std   0.125487    0.014388                 98807.812136   \n",
+       "mean  0.958916    0.086565               -520945.104332   \n",
+       "std   0.314371    0.049576                 80345.848916   \n",
        "\n",
-       "      test_neg_mean_absolute_error   test_r2           model_name  \n",
-       "mean                   -521.421224  0.867283  K-Nearest Neighbors  \n",
-       "std                      56.917455  0.024541  K-Nearest Neighbors  "
+       "      test_neg_mean_absolute_error   test_r2  \n",
+       "mean                   -538.463882  0.859563  \n",
+       "std                      36.727944  0.024787  "
       ]
      },
      "metadata": {},
@@ -642,27 +998,24 @@
        "      <th>test_neg_mean_squared_error</th>\n",
        "      <th>test_neg_mean_absolute_error</th>\n",
        "      <th>test_r2</th>\n",
-       "      <th>model_name</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>mean</th>\n",
-       "      <td>42.909599</td>\n",
-       "      <td>0.016341</td>\n",
-       "      <td>-11775.561442</td>\n",
-       "      <td>-68.763639</td>\n",
-       "      <td>0.996826</td>\n",
-       "      <td>Gradient Boosting</td>\n",
+       "      <td>87.815198</td>\n",
+       "      <td>0.043461</td>\n",
+       "      <td>-11767.095009</td>\n",
+       "      <td>-69.842862</td>\n",
+       "      <td>0.996832</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>std</th>\n",
-       "      <td>8.347175</td>\n",
-       "      <td>0.001307</td>\n",
-       "      <td>3159.517240</td>\n",
-       "      <td>7.482689</td>\n",
-       "      <td>0.000856</td>\n",
-       "      <td>Gradient Boosting</td>\n",
+       "      <td>15.660049</td>\n",
+       "      <td>0.012917</td>\n",
+       "      <td>2630.227186</td>\n",
+       "      <td>6.533291</td>\n",
+       "      <td>0.000709</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -670,12 +1023,12 @@
       ],
       "text/plain": [
        "       fit_time  score_time  test_neg_mean_squared_error  \\\n",
-       "mean  42.909599    0.016341                -11775.561442   \n",
-       "std    8.347175    0.001307                  3159.517240   \n",
+       "mean  87.815198    0.043461                -11767.095009   \n",
+       "std   15.660049    0.012917                  2630.227186   \n",
        "\n",
-       "      test_neg_mean_absolute_error   test_r2         model_name  \n",
-       "mean                    -68.763639  0.996826  Gradient Boosting  \n",
-       "std                       7.482689  0.000856  Gradient Boosting  "
+       "      test_neg_mean_absolute_error   test_r2  \n",
+       "mean                    -69.842862  0.996832  \n",
+       "std                       6.533291  0.000709  "
       ]
      },
      "metadata": {},
@@ -714,27 +1067,24 @@
        "      <th>test_neg_mean_squared_error</th>\n",
        "      <th>test_neg_mean_absolute_error</th>\n",
        "      <th>test_r2</th>\n",
-       "      <th>model_name</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>mean</th>\n",
-       "      <td>0.684064</td>\n",
-       "      <td>0.018473</td>\n",
-       "      <td>-59390.980641</td>\n",
-       "      <td>-164.245985</td>\n",
-       "      <td>0.983967</td>\n",
-       "      <td>Decision Tree</td>\n",
+       "      <td>1.580274</td>\n",
+       "      <td>0.043137</td>\n",
+       "      <td>-57611.119230</td>\n",
+       "      <td>-163.296206</td>\n",
+       "      <td>0.984511</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>std</th>\n",
-       "      <td>0.106449</td>\n",
-       "      <td>0.006129</td>\n",
-       "      <td>12130.331634</td>\n",
-       "      <td>17.579599</td>\n",
-       "      <td>0.003618</td>\n",
-       "      <td>Decision Tree</td>\n",
+       "      <td>0.257548</td>\n",
+       "      <td>0.014953</td>\n",
+       "      <td>10015.212231</td>\n",
+       "      <td>14.241930</td>\n",
+       "      <td>0.002719</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -742,12 +1092,12 @@
       ],
       "text/plain": [
        "      fit_time  score_time  test_neg_mean_squared_error  \\\n",
-       "mean  0.684064    0.018473                -59390.980641   \n",
-       "std   0.106449    0.006129                 12130.331634   \n",
+       "mean  1.580274    0.043137                -57611.119230   \n",
+       "std   0.257548    0.014953                 10015.212231   \n",
        "\n",
-       "      test_neg_mean_absolute_error   test_r2     model_name  \n",
-       "mean                   -164.245985  0.983967  Decision Tree  \n",
-       "std                      17.579599  0.003618  Decision Tree  "
+       "      test_neg_mean_absolute_error   test_r2  \n",
+       "mean                   -163.296206  0.984511  \n",
+       "std                      14.241930  0.002719  "
       ]
      },
      "metadata": {},
@@ -786,27 +1136,24 @@
        "      <th>test_neg_mean_squared_error</th>\n",
        "      <th>test_neg_mean_absolute_error</th>\n",
        "      <th>test_r2</th>\n",
-       "      <th>model_name</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>mean</th>\n",
-       "      <td>10.112643</td>\n",
-       "      <td>0.025550</td>\n",
-       "      <td>-18307.274898</td>\n",
-       "      <td>-83.639403</td>\n",
-       "      <td>0.995059</td>\n",
-       "      <td>Random Forest</td>\n",
+       "      <td>25.657500</td>\n",
+       "      <td>0.046849</td>\n",
+       "      <td>-18545.332476</td>\n",
+       "      <td>-85.850864</td>\n",
+       "      <td>0.995005</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>std</th>\n",
-       "      <td>1.794122</td>\n",
-       "      <td>0.014259</td>\n",
-       "      <td>5609.330859</td>\n",
-       "      <td>11.129464</td>\n",
-       "      <td>0.001523</td>\n",
-       "      <td>Random Forest</td>\n",
+       "      <td>5.260464</td>\n",
+       "      <td>0.015731</td>\n",
+       "      <td>5785.896650</td>\n",
+       "      <td>10.240212</td>\n",
+       "      <td>0.001532</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -814,12 +1161,12 @@
       ],
       "text/plain": [
        "       fit_time  score_time  test_neg_mean_squared_error  \\\n",
-       "mean  10.112643    0.025550                -18307.274898   \n",
-       "std    1.794122    0.014259                  5609.330859   \n",
+       "mean  25.657500    0.046849                -18545.332476   \n",
+       "std    5.260464    0.015731                  5785.896650   \n",
        "\n",
-       "      test_neg_mean_absolute_error   test_r2     model_name  \n",
-       "mean                    -83.639403  0.995059  Random Forest  \n",
-       "std                      11.129464  0.001523  Random Forest  "
+       "      test_neg_mean_absolute_error   test_r2  \n",
+       "mean                    -85.850864  0.995005  \n",
+       "std                      10.240212  0.001532  "
       ]
      },
      "metadata": {},
@@ -827,6 +1174,7 @@
     }
    ],
    "source": [
+    "model_results = {}\n",
     "results = pd.DataFrame({})\n",
     "cross_validate_grid_search = KFold(n_splits=n_folds_grid_search)\n",
     "cross_validate_comparative_analysis = ShuffleSplit(n_splits=n_splits_comparative_analysis, test_size=test_size, random_state=random_state)\n",
@@ -846,6 +1194,20 @@
     "        ('model', model_grid_search)\n",
     "    ])\n",
     "    \n",
+    "    # Armazenando os resultados de y_true e y_pred\n",
+    "    for train_index, test_index in cross_validate_comparative_analysis.split(X, y):\n",
+    "        X_train, X_test = X.iloc[train_index], X.iloc[test_index]\n",
+    "        y_train, y_test = y.iloc[train_index], y.iloc[test_index]\n",
+    "        \n",
+    "        approach.fit(X_train, y_train)\n",
+    "        y_pred = approach.predict(X_test)\n",
+    "        \n",
+    "        model_results[model_name] = {\n",
+    "            'y_true': y_test,\n",
+    "            'y_pred': y_pred\n",
+    "        }\n",
+    "\n",
+    "    # Avaliando o modelo com cross_validate\n",
     "    scores = cross_validate(\n",
     "        estimator=approach,\n",
     "        X=X,\n",
@@ -856,15 +1218,16 @@
     "    )\n",
     "    \n",
     "    scores_df = pd.DataFrame(scores)\n",
-    "    aggregated_scores = scores_df.agg(['mean', 'std'])\n",
-    "    aggregated_scores['model_name'] = model_name\n",
-    "    display(aggregated_scores)\n",
-    "    results = pd.concat([results, aggregated_scores], ignore_index=True)"
+    "    scores_df['model_name'] = model_name\n",
+    "    results = pd.concat([results, scores_df], ignore_index=True)\n",
+    "    numeric_scores_df = scores_df.select_dtypes(include=['float64', 'int64'])\n",
+    "    scores_aggregated = numeric_scores_df.agg(['mean', 'std'])\n",
+    "    display(scores_aggregated)"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 14,
    "id": "5127adfc-d0c6-4f15-b833-bf1fdc91ef1b",
    "metadata": {
     "tags": []
@@ -873,7 +1236,7 @@
     {
      "data": {
       "text/markdown": [
-       "### 3.1 Resultados gerais"
+       "### 2.1 Resultados gerais"
       ],
       "text/plain": [
        "<IPython.core.display.Markdown object>"
@@ -886,110 +1249,110 @@
      "data": {
       "text/html": [
        "<style type=\"text/css\">\n",
-       "#T_c6828 td {\n",
+       "#T_6f080 td {\n",
        "  text-align: center;\n",
        "}\n",
-       "#T_c6828_row0_col0, #T_c6828_row2_col1, #T_c6828_row4_col1, #T_c6828_row6_col1, #T_c6828_row8_col1 {\n",
+       "#T_6f080_row0_col2, #T_6f080_row2_col0, #T_6f080_row4_col1, #T_6f080_row6_col1, #T_6f080_row8_col1 {\n",
        "  color: white;\n",
        "  background-color: gray;\n",
        "  font-weight: bold;\n",
        "}\n",
        "</style>\n",
-       "<table id=\"T_c6828\">\n",
+       "<table id=\"T_6f080\">\n",
        "  <thead>\n",
        "    <tr>\n",
        "      <th class=\"blank\" >&nbsp;</th>\n",
        "      <th class=\"index_name level0\" >model_name</th>\n",
-       "      <th id=\"T_c6828_level0_col0\" class=\"col_heading level0 col0\" >Decision Tree</th>\n",
-       "      <th id=\"T_c6828_level0_col1\" class=\"col_heading level0 col1\" >Gradient Boosting</th>\n",
-       "      <th id=\"T_c6828_level0_col2\" class=\"col_heading level0 col2\" >K-Nearest Neighbors</th>\n",
-       "      <th id=\"T_c6828_level0_col3\" class=\"col_heading level0 col3\" >Random Forest</th>\n",
+       "      <th id=\"T_6f080_level0_col0\" class=\"col_heading level0 col0\" >Decision Tree</th>\n",
+       "      <th id=\"T_6f080_level0_col1\" class=\"col_heading level0 col1\" >Gradient Boosting</th>\n",
+       "      <th id=\"T_6f080_level0_col2\" class=\"col_heading level0 col2\" >K-Nearest Neighbors</th>\n",
+       "      <th id=\"T_6f080_level0_col3\" class=\"col_heading level0 col3\" >Random Forest</th>\n",
        "    </tr>\n",
        "  </thead>\n",
        "  <tbody>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level0_row0\" class=\"row_heading level0 row0\" rowspan=\"2\">fit_time</th>\n",
-       "      <th id=\"T_c6828_level1_row0\" class=\"row_heading level1 row0\" >mean</th>\n",
-       "      <td id=\"T_c6828_row0_col0\" class=\"data row0 col0\" >0.395256</td>\n",
-       "      <td id=\"T_c6828_row0_col1\" class=\"data row0 col1\" >25.628387</td>\n",
-       "      <td id=\"T_c6828_row0_col2\" class=\"data row0 col2\" >0.451770</td>\n",
-       "      <td id=\"T_c6828_row0_col3\" class=\"data row0 col3\" >5.953382</td>\n",
+       "      <th id=\"T_6f080_level0_row0\" class=\"row_heading level0 row0\" rowspan=\"2\">fit_time</th>\n",
+       "      <th id=\"T_6f080_level1_row0\" class=\"row_heading level1 row0\" >mean</th>\n",
+       "      <td id=\"T_6f080_row0_col0\" class=\"data row0 col0\" >1.580274</td>\n",
+       "      <td id=\"T_6f080_row0_col1\" class=\"data row0 col1\" >87.815198</td>\n",
+       "      <td id=\"T_6f080_row0_col2\" class=\"data row0 col2\" >0.958916</td>\n",
+       "      <td id=\"T_6f080_row0_col3\" class=\"data row0 col3\" >25.657500</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level1_row1\" class=\"row_heading level1 row1\" >std</th>\n",
-       "      <td id=\"T_c6828_row1_col0\" class=\"data row1 col0\" >0.408435</td>\n",
-       "      <td id=\"T_c6828_row1_col1\" class=\"data row1 col1\" >24.439324</td>\n",
-       "      <td id=\"T_c6828_row1_col2\" class=\"data row1 col2\" >0.461433</td>\n",
-       "      <td id=\"T_c6828_row1_col3\" class=\"data row1 col3\" >5.882082</td>\n",
+       "      <th id=\"T_6f080_level1_row1\" class=\"row_heading level1 row1\" >std</th>\n",
+       "      <td id=\"T_6f080_row1_col0\" class=\"data row1 col0\" >0.257548</td>\n",
+       "      <td id=\"T_6f080_row1_col1\" class=\"data row1 col1\" >15.660049</td>\n",
+       "      <td id=\"T_6f080_row1_col2\" class=\"data row1 col2\" >0.314371</td>\n",
+       "      <td id=\"T_6f080_row1_col3\" class=\"data row1 col3\" >5.260464</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level0_row2\" class=\"row_heading level0 row2\" rowspan=\"2\">score_time</th>\n",
-       "      <th id=\"T_c6828_level1_row2\" class=\"row_heading level1 row2\" >mean</th>\n",
-       "      <td id=\"T_c6828_row2_col0\" class=\"data row2 col0\" >0.012301</td>\n",
-       "      <td id=\"T_c6828_row2_col1\" class=\"data row2 col1\" >0.008824</td>\n",
-       "      <td id=\"T_c6828_row2_col2\" class=\"data row2 col2\" >0.022127</td>\n",
-       "      <td id=\"T_c6828_row2_col3\" class=\"data row2 col3\" >0.019905</td>\n",
+       "      <th id=\"T_6f080_level0_row2\" class=\"row_heading level0 row2\" rowspan=\"2\">score_time</th>\n",
+       "      <th id=\"T_6f080_level1_row2\" class=\"row_heading level1 row2\" >mean</th>\n",
+       "      <td id=\"T_6f080_row2_col0\" class=\"data row2 col0\" >0.043137</td>\n",
+       "      <td id=\"T_6f080_row2_col1\" class=\"data row2 col1\" >0.043461</td>\n",
+       "      <td id=\"T_6f080_row2_col2\" class=\"data row2 col2\" >0.086565</td>\n",
+       "      <td id=\"T_6f080_row2_col3\" class=\"data row2 col3\" >0.046849</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level1_row3\" class=\"row_heading level1 row3\" >std</th>\n",
-       "      <td id=\"T_c6828_row3_col0\" class=\"data row3 col0\" >0.008728</td>\n",
-       "      <td id=\"T_c6828_row3_col1\" class=\"data row3 col1\" >0.010631</td>\n",
-       "      <td id=\"T_c6828_row3_col2\" class=\"data row3 col2\" >0.010944</td>\n",
-       "      <td id=\"T_c6828_row3_col3\" class=\"data row3 col3\" >0.007984</td>\n",
+       "      <th id=\"T_6f080_level1_row3\" class=\"row_heading level1 row3\" >std</th>\n",
+       "      <td id=\"T_6f080_row3_col0\" class=\"data row3 col0\" >0.014953</td>\n",
+       "      <td id=\"T_6f080_row3_col1\" class=\"data row3 col1\" >0.012917</td>\n",
+       "      <td id=\"T_6f080_row3_col2\" class=\"data row3 col2\" >0.049576</td>\n",
+       "      <td id=\"T_6f080_row3_col3\" class=\"data row3 col3\" >0.015731</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level0_row4\" class=\"row_heading level0 row4\" rowspan=\"2\">test_neg_mean_squared_error</th>\n",
-       "      <th id=\"T_c6828_level1_row4\" class=\"row_heading level1 row4\" >mean</th>\n",
-       "      <td id=\"T_c6828_row4_col0\" class=\"data row4 col0\" >-23630.324504</td>\n",
-       "      <td id=\"T_c6828_row4_col1\" class=\"data row4 col1\" >-4308.022101</td>\n",
-       "      <td id=\"T_c6828_row4_col2\" class=\"data row4 col2\" >-198138.736726</td>\n",
-       "      <td id=\"T_c6828_row4_col3\" class=\"data row4 col3\" >-6348.972019</td>\n",
+       "      <th id=\"T_6f080_level0_row4\" class=\"row_heading level0 row4\" rowspan=\"2\">test_neg_mean_squared_error</th>\n",
+       "      <th id=\"T_6f080_level1_row4\" class=\"row_heading level1 row4\" >mean</th>\n",
+       "      <td id=\"T_6f080_row4_col0\" class=\"data row4 col0\" >-57611.119230</td>\n",
+       "      <td id=\"T_6f080_row4_col1\" class=\"data row4 col1\" >-11767.095009</td>\n",
+       "      <td id=\"T_6f080_row4_col2\" class=\"data row4 col2\" >-520945.104332</td>\n",
+       "      <td id=\"T_6f080_row4_col3\" class=\"data row4 col3\" >-18545.332476</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level1_row5\" class=\"row_heading level1 row5\" >std</th>\n",
-       "      <td id=\"T_c6828_row5_col0\" class=\"data row5 col0\" >50573.204909</td>\n",
-       "      <td id=\"T_c6828_row5_col1\" class=\"data row5 col1\" >10560.695413</td>\n",
-       "      <td id=\"T_c6828_row5_col2\" class=\"data row5 col2\" >419945.836699</td>\n",
-       "      <td id=\"T_c6828_row5_col3\" class=\"data row5 col3\" >16911.594114</td>\n",
+       "      <th id=\"T_6f080_level1_row5\" class=\"row_heading level1 row5\" >std</th>\n",
+       "      <td id=\"T_6f080_row5_col0\" class=\"data row5 col0\" >10015.212231</td>\n",
+       "      <td id=\"T_6f080_row5_col1\" class=\"data row5 col1\" >2630.227186</td>\n",
+       "      <td id=\"T_6f080_row5_col2\" class=\"data row5 col2\" >80345.848916</td>\n",
+       "      <td id=\"T_6f080_row5_col3\" class=\"data row5 col3\" >5785.896650</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level0_row6\" class=\"row_heading level0 row6\" rowspan=\"2\">test_neg_mean_absolute_error</th>\n",
-       "      <th id=\"T_c6828_level1_row6\" class=\"row_heading level1 row6\" >mean</th>\n",
-       "      <td id=\"T_c6828_row6_col0\" class=\"data row6 col0\" >-73.333193</td>\n",
-       "      <td id=\"T_c6828_row6_col1\" class=\"data row6 col1\" >-30.640475</td>\n",
-       "      <td id=\"T_c6828_row6_col2\" class=\"data row6 col2\" >-232.251885</td>\n",
-       "      <td id=\"T_c6828_row6_col3\" class=\"data row6 col3\" >-36.254970</td>\n",
+       "      <th id=\"T_6f080_level0_row6\" class=\"row_heading level0 row6\" rowspan=\"2\">test_neg_mean_absolute_error</th>\n",
+       "      <th id=\"T_6f080_level1_row6\" class=\"row_heading level1 row6\" >mean</th>\n",
+       "      <td id=\"T_6f080_row6_col0\" class=\"data row6 col0\" >-163.296206</td>\n",
+       "      <td id=\"T_6f080_row6_col1\" class=\"data row6 col1\" >-69.842862</td>\n",
+       "      <td id=\"T_6f080_row6_col2\" class=\"data row6 col2\" >-538.463882</td>\n",
+       "      <td id=\"T_6f080_row6_col3\" class=\"data row6 col3\" >-85.850864</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level1_row7\" class=\"row_heading level1 row7\" >std</th>\n",
-       "      <td id=\"T_c6828_row7_col0\" class=\"data row7 col0\" >128.570104</td>\n",
-       "      <td id=\"T_c6828_row7_col1\" class=\"data row7 col1\" >53.914296</td>\n",
-       "      <td id=\"T_c6828_row7_col2\" class=\"data row7 col2\" >408.947202</td>\n",
-       "      <td id=\"T_c6828_row7_col3\" class=\"data row7 col3\" >67.011709</td>\n",
+       "      <th id=\"T_6f080_level1_row7\" class=\"row_heading level1 row7\" >std</th>\n",
+       "      <td id=\"T_6f080_row7_col0\" class=\"data row7 col0\" >14.241930</td>\n",
+       "      <td id=\"T_6f080_row7_col1\" class=\"data row7 col1\" >6.533291</td>\n",
+       "      <td id=\"T_6f080_row7_col2\" class=\"data row7 col2\" >36.727944</td>\n",
+       "      <td id=\"T_6f080_row7_col3\" class=\"data row7 col3\" >10.240212</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level0_row8\" class=\"row_heading level0 row8\" rowspan=\"2\">test_r2</th>\n",
-       "      <th id=\"T_c6828_level1_row8\" class=\"row_heading level1 row8\" >mean</th>\n",
-       "      <td id=\"T_c6828_row8_col0\" class=\"data row8 col0\" >0.493793</td>\n",
-       "      <td id=\"T_c6828_row8_col1\" class=\"data row8 col1\" >0.498841</td>\n",
-       "      <td id=\"T_c6828_row8_col2\" class=\"data row8 col2\" >0.445912</td>\n",
-       "      <td id=\"T_c6828_row8_col3\" class=\"data row8 col3\" >0.498291</td>\n",
+       "      <th id=\"T_6f080_level0_row8\" class=\"row_heading level0 row8\" rowspan=\"2\">test_r2</th>\n",
+       "      <th id=\"T_6f080_level1_row8\" class=\"row_heading level1 row8\" >mean</th>\n",
+       "      <td id=\"T_6f080_row8_col0\" class=\"data row8 col0\" >0.984511</td>\n",
+       "      <td id=\"T_6f080_row8_col1\" class=\"data row8 col1\" >0.996832</td>\n",
+       "      <td id=\"T_6f080_row8_col2\" class=\"data row8 col2\" >0.859563</td>\n",
+       "      <td id=\"T_6f080_row8_col3\" class=\"data row8 col3\" >0.995005</td>\n",
        "    </tr>\n",
        "    <tr>\n",
-       "      <th id=\"T_c6828_level1_row9\" class=\"row_heading level1 row9\" >std</th>\n",
-       "      <td id=\"T_c6828_row9_col0\" class=\"data row9 col0\" >0.693211</td>\n",
-       "      <td id=\"T_c6828_row9_col1\" class=\"data row9 col1\" >0.704257</td>\n",
-       "      <td id=\"T_c6828_row9_col2\" class=\"data row9 col2\" >0.595908</td>\n",
-       "      <td id=\"T_c6828_row9_col3\" class=\"data row9 col3\" >0.702536</td>\n",
+       "      <th id=\"T_6f080_level1_row9\" class=\"row_heading level1 row9\" >std</th>\n",
+       "      <td id=\"T_6f080_row9_col0\" class=\"data row9 col0\" >0.002719</td>\n",
+       "      <td id=\"T_6f080_row9_col1\" class=\"data row9 col1\" >0.000709</td>\n",
+       "      <td id=\"T_6f080_row9_col2\" class=\"data row9 col2\" >0.024787</td>\n",
+       "      <td id=\"T_6f080_row9_col3\" class=\"data row9 col3\" >0.001532</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n"
       ],
       "text/plain": [
-       "<pandas.io.formats.style.Styler at 0x7fc05d2c0a30>"
+       "<pandas.io.formats.style.Styler at 0x7f79295aece0>"
       ]
      },
-     "execution_count": 10,
+     "execution_count": 14,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1001,7 +1364,7 @@
     "            return np.where(s == np.nanmin(s.values), props, '')\n",
     "        return np.where(s == np.nanmax(s.values), props, '')\n",
     "\n",
-    "display(Markdown(\"### 3.1 Resultados gerais\"))\n",
+    "display(Markdown(\"### 2.1 Resultados gerais\"))\n",
     "(\n",
     "    results\n",
     "    .groupby('model_name')\n",
@@ -1025,17 +1388,52 @@
     "Esses resultados sugerem que, se o objetivo é maximizar a precisão e há tempo computacional disponível, o Gradient Boosting é a melhor opção. Se o tempo de treinamento for uma restrição, o Random Forest pode ser uma boa alternativa."
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "id": "38a47820-08a4-4bd3-b8a8-88a8a8ce31d8",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "image/png": "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",
+      "text/plain": [
+       "<Figure size 1400x1200 with 4 Axes>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "fig, axs = plt.subplots(2, 2, figsize=(14, 12))\n",
+    "\n",
+    "for i, (model_name, result) in enumerate(model_results.items()):\n",
+    "    row = i // 2  # Determina a linha\n",
+    "    col = i % 2   # Determina a coluna\n",
+    "    \n",
+    "    axs[row, col].scatter(result['y_true'], result['y_pred'] - result['y_true'])\n",
+    "    axs[row, col].hlines(y=0, xmin=result['y_true'].min(), xmax=result['y_true'].max(), colors='r')\n",
+    "    axs[row, col].set_xlabel('True Values')\n",
+    "    axs[row, col].set_ylabel('Residuals')\n",
+    "    axs[row, col].set_title(f'Residuals Plot for {model_name}')\n",
+    "\n",
+    "\n",
+    "plt.tight_layout()\n",
+    "plt.show()"
+   ]
+  },
   {
    "cell_type": "markdown",
    "id": "71d8f5fb-77b4-4f65-9cde-6ab2ec8fc40a",
    "metadata": {},
    "source": [
-    "## 3.2 Persistência do modelo"
+    "### 2.2 Persistência do modelo"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 17,
    "id": "2003213b-5432-4d3d-b6ba-ab7377c518f3",
    "metadata": {
     "tags": []
@@ -1045,7 +1443,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Hiper parâmetros do modelo: {'learning_rate': 0.2, 'max_depth': 3, 'n_estimators': 150}\n"
+      "Hiper parâmetros do modelo: {'learning_rate': 0.1, 'max_depth': 3, 'n_estimators': 150}\n"
      ]
     }
    ],
@@ -1072,7 +1470,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 18,
    "id": "c9868cb2-a2c8-4963-9535-04667f54f742",
    "metadata": {
     "tags": []
@@ -1084,7 +1482,7 @@
        "['../models/model.joblib']"
       ]
      },
-     "execution_count": 12,
+     "execution_count": 18,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1110,7 +1508,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.12"
+   "version": "3.10.11"
   }
  },
  "nbformat": 4,