arthur/ArtStudies
1
0
Fork 0
mirror of https://github.com/ArthurDanjou/ArtStudies.git synced 2026-01-27 06:54:18 +01:00
Code Issues Packages Projects Releases Wiki Activity

Refactor code for improved readability and consistency across notebooks

Browse Source 
- Standardized spacing around operators and function arguments in TP7_Kmeans.ipynb and neural_network.ipynb.
- Enhanced the formatting of model building and training code in neural_network.ipynb for better clarity.
- Updated the pyproject.toml to remove a specific TensorFlow version and added linting configuration for Ruff.
- Improved comments and organization in the code to facilitate easier understanding and maintenance.
This commit is contained in:
Arthur Danjou
2025-07-01 20:46:08 +02:00
parent e273cf90f7
commit f94ff07cab
34 changed files with 5713 additions and 5047 deletions
Download Patch File Download Diff File Expand all files Collapse all files

83
M1/Statistical Learning/neural_network.ipynb
View File

@@ -6,7 +6,7 @@
"metadata": {},
"outputs": [],
"source": [
"import numpy as np \n",
"import numpy as np\n",
"import pandas as pd\n",
"import tensorflow as tf\n",
"import matplotlib.pyplot as plt"
@@ -178,16 +178,21 @@
"outputs": [],
"source": [
"def build_model():\n",
" model = tf.keras.models.Sequential([\n",
" tf.keras.layers.Dense(16, activation='relu', input_shape=(X.shape[1],), kernel_regularizer=tf.keras.regularizers.l2(0.01)),\n",
" tf.keras.layers.Dense(8, activation='relu', kernel_regularizer=tf.keras.regularizers.l2(0.01)),\n",
" tf.keras.layers.Dense(1, activation='sigmoid')\n",
" ])\n",
" model.compile(\n",
" optimizer='adam',\n",
" loss='binary_crossentropy',\n",
" metrics=['accuracy']\n",
" model = tf.keras.models.Sequential(\n",
" [\n",
" tf.keras.layers.Dense(\n",
" 16,\n",
" activation=\"relu\",\n",
" input_shape=(X.shape[1],),\n",
" kernel_regularizer=tf.keras.regularizers.l2(0.01),\n",
" ),\n",
" tf.keras.layers.Dense(\n",
" 8, activation=\"relu\", kernel_regularizer=tf.keras.regularizers.l2(0.01)\n",
" ),\n",
" tf.keras.layers.Dense(1, activation=\"sigmoid\"),\n",
" ]\n",
" )\n",
" model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"accuracy\"])\n",
" return model"
]
},
@@ -291,10 +296,7 @@
"histories = []\n",
"\n",
"early_stopping = EarlyStopping(\n",
" monitor='val_loss',\n",
" patience=10,\n",
" restore_best_weights=True,\n",
" verbose=1\n",
" monitor=\"val_loss\", patience=10, restore_best_weights=True, verbose=1\n",
")\n",
"\n",
"for fold, (train_idx, val_idx) in enumerate(skf.split(X, y), 1):\n",
@@ -305,29 +307,28 @@
" scaler = StandardScaler()\n",
" X_train_scaled = scaler.fit_transform(X_train)\n",
" X_val_scaled = scaler.transform(X_val)\n",
" \n",
"\n",
" model = build_model()\n",
"\n",
" model.compile(\n",
" optimizer='adam',\n",
" loss='binary_crossentropy',\n",
" metrics=[\"f1_score\"]\n",
" )\n",
" model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"f1_score\"])\n",
"\n",
" # EarlyStopping\n",
" callback = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n",
" callback = tf.keras.callbacks.EarlyStopping(\n",
" monitor=\"val_loss\", patience=10, restore_best_weights=True\n",
" )\n",
"\n",
" # Entraînement\n",
" history = model.fit(\n",
" X_train_scaled, y_train,\n",
" X_train_scaled,\n",
" y_train,\n",
" epochs=50,\n",
" batch_size=8,\n",
" validation_data=(X_val_scaled, y_val),\n",
" callbacks=[callback],\n",
" verbose=0,\n",
" class_weight={0: 1.0, 1: 2.0}\n",
" class_weight={0: 1.0, 1: 2.0},\n",
" )\n",
" \n",
"\n",
" histories.append(history.history)\n",
"\n",
" # Prédiction & F1\n",
@@ -360,9 +361,9 @@
"axes = axes.flatten() # Flatten to easily iterate\n",
"\n",
"for i, (hist, ax) in enumerate(zip(histories, axes)):\n",
" ax.plot(hist['loss'], label='Train loss', alpha=0.6)\n",
" ax.plot(hist['val_loss'], label='Val loss', linestyle='--', alpha=0.6)\n",
" ax.set_title(f\"Fold {i+1}\")\n",
" ax.plot(hist[\"loss\"], label=\"Train loss\", alpha=0.6)\n",
" ax.plot(hist[\"val_loss\"], label=\"Val loss\", linestyle=\"--\", alpha=0.6)\n",
" ax.set_title(f\"Fold {i + 1}\")\n",
" ax.set_xlabel(\"Epochs\")\n",
" if i % 2 == 0:\n",
" ax.set_ylabel(\"Binary Crossentropy\")\n",
@@ -436,7 +437,9 @@
"import tensorflow as tf\n",
"import numpy as np\n",
"\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify=y)\n",
"X_train, X_test, y_train, y_test = train_test_split(\n",
" X, y, test_size=0.2, random_state=42, stratify=y\n",
")\n",
"\n",
"scaler = StandardScaler()\n",
"X_train_scaled = scaler.fit_transform(X_train)\n",
@@ -444,21 +447,21 @@
"\n",
"model = build_model()\n",
"\n",
"model.compile(\n",
" optimizer='adam',\n",
" loss='binary_crossentropy'\n",
"model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\")\n",
"\n",
"callback = tf.keras.callbacks.EarlyStopping(\n",
" monitor=\"val_loss\", patience=10, restore_best_weights=True\n",
")\n",
"\n",
"callback = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n",
"\n",
"history = model.fit(\n",
" X_train_scaled, y_train,\n",
" X_train_scaled,\n",
" y_train,\n",
" epochs=50,\n",
" batch_size=8,\n",
" validation_split=0.2,\n",
" callbacks=[callback],\n",
" verbose=0,\n",
" class_weight={0: 1.0, 1: 2.0}\n",
" class_weight={0: 1.0, 1: 2.0},\n",
")\n",
"\n",
"\n",
@@ -486,11 +489,11 @@
],
"source": [
"plt.figure(figsize=(8, 5))\n",
"plt.plot(history.history['loss'], label='Loss (train)')\n",
"plt.plot(history.history['val_loss'], label='Loss (val)', linestyle='--')\n",
"plt.xlabel('Epochs')\n",
"plt.ylabel('Binary Cross-Entropy Loss')\n",
"plt.title('Courbe d\\'apprentissage')\n",
"plt.plot(history.history[\"loss\"], label=\"Loss (train)\")\n",
"plt.plot(history.history[\"val_loss\"], label=\"Loss (val)\", linestyle=\"--\")\n",
"plt.xlabel(\"Epochs\")\n",
"plt.ylabel(\"Binary Cross-Entropy Loss\")\n",
"plt.title(\"Courbe d'apprentissage\")\n",
"plt.legend()\n",
"plt.grid(True)\n",
"plt.tight_layout()\n",