"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"corr = processed_train.drop([\"ID\",\"URL\"], axis=1)\n",
"\n",
"import seaborn as sns\n",
"\n",
"plt.figure(figsize=(20, 8))\n",
"sns.heatmap(corr.corr(), annot=True, cmap='coolwarm')\n",
"plt.title('Correlation Heatmap')\n",
"plt.show()\n"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {},
"outputs": [],
"source": [
"X = processed_train.drop([\"ID\", \"URL\",\"label\"], axis=1)\n",
"X = X.astype(np.float32)\n"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
" feature VIF\n",
"0 url_length_cat 5.411857\n",
"1 num_dots 2.960486\n",
"2 num_digits 3.312505\n",
"3 num_special_chars 1.474281\n",
"4 url_keyword 1.201418\n",
"5 num_underbar 1.534030\n",
"6 extract_consecutive_numbers 1.814039\n",
"7 number 2.817060\n",
"8 upper 1.307657\n",
"9 is_common_tld 1.657451\n",
"10 is_country_tld 1.433461\n",
"11 is_suspicious_tld 1.104899\n",
"12 domain_length 2.359749\n",
"13 has_subdomain 3.681469\n",
"14 subdomain_length 3.593552\n",
"15 subdomain_count 6.335112\n",
"16 path_depth 2.027944\n",
"17 has_query 2.460386\n",
"18 query_length 3.318923\n",
"19 query_param_count 3.128829\n",
"20 url_shorteners 1.036828\n",
"21 compression_ratio 5.641054\n",
"22 check_similar_brand 1.001343\n",
"23 entropy 4.319254\n",
"24 digit_ratio 3.393270\n",
"25 special_char_ratio 1.755469\n"
]
}
],
"source": [
"from statsmodels.stats.outliers_influence import variance_inflation_factor\n",
"from statsmodels.tools.tools import add_constant\n",
"import statsmodels.api as sm # statsmodels를 임포트\n",
"\n",
"# 상수항 추가\n",
"X_with_const = add_constant(X)\n",
"\n",
"# VIF 계산\n",
"vif_data = pd.DataFrame()\n",
"vif_data[\"feature\"] = X.columns\n",
"vif_data[\"VIF\"] = [variance_inflation_factor(X_with_const.values, i+1) for i in range(len(X.columns))]\n",
"\n",
"print(vif_data)"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [],
"source": [
"X = processed_train.drop([\"ID\",\"URL\", \"label\"], axis=1)\n",
"y = processed_train[\"label\"]"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.model_selection import train_test_split\n",
"\n",
"# 1차 분할\n",
"X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42, stratify = y)\n"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {},
"outputs": [],
"source": [
"# from imblearn.pipeline import Pipeline\n",
"# from imblearn.under_sampling import RandomUnderSampler\n",
"# from imblearn.over_sampling import SMOTE\n",
"\n",
"# # 언더샘플링 - 다수 클래스(정상) 60%로 줄임\n",
"# under = RandomUnderSampler(sampling_strategy=0.6, random_state=42)\n",
"\n",
"# # 2단계: 오버샘플링 - 소수 클래스(악성)를 다수 클래스 수에 맞게 증강\n",
"# over = SMOTE(sampling_strategy='auto', random_state=42)\n",
"\n",
"# # 파이프라인 구성\n",
"# hybrid_sampler = Pipeline(steps=[\n",
"# ('undersample', under),\n",
"# ('oversample', over)\n",
"# ])\n",
"\n",
"# # 샘플링 적용\n",
"# X_train, y_train = hybrid_sampler.fit_resample(X_train, y_train)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [],
"source": [
"from sklearn.preprocessing import MinMaxScaler\n",
"import pickle\n",
"scaler = MinMaxScaler()\n",
"X_train_scaled = scaler.fit_transform(X_train)\n",
"X_test_scaled = scaler.transform(X_test)\n",
"\n",
"\n",
"with open('scaler.pkl','wb') as f:\n",
" pickle.dump(scaler, f)\n"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [],
"source": [
"import tensorflow as tf\n",
"from tensorflow.keras.layers import Dense, Dropout, BatchNormalization\n",
"\n",
"def build_model(input_dim, learning_rate=0.001):\n",
" \"\"\"\n",
" 딥러닝 모델 생성 및 컴파일\n",
"\n",
" Parameters:\n",
" input_dim (int): 입력 특성 수 (X.shape[1])\n",
" learning_rate (float): 옵티마이저 학습률\n",
"\n",
" Returns:\n",
" tf.keras.Model: 컴파일된 Keras Sequential 모델\n",
" \"\"\"\n",
" model = tf.keras.Sequential([\n",
" tf.keras.layers.Input(shape=(input_dim,)),\n",
" Dense(128, activation='relu'),\n",
" Dense(64, activation='relu'),\n",
" Dense(1, activation='sigmoid')\n",
" ], name=\"Model\")\n",
"\n",
" model.compile(\n",
" optimizer=tf.keras.optimizers.Adam(learning_rate=learning_rate),\n",
" loss='binary_crossentropy',\n",
" metrics=['accuracy']\n",
" )\n",
"\n",
" return model\n"
]
},
{
"cell_type": "code",
"execution_count": 19,
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"Model: \"Model\"\n",
"\n"
],
"text/plain": [
"\u001b[1mModel: \"Model\"\u001b[0m\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃ Layer (type) ┃ Output Shape ┃ Param # ┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ dense (Dense) │ (None, 128) │ 3,456 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (Dense) │ (None, 64) │ 8,256 │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (Dense) │ (None, 1) │ 65 │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
"\n"
],
"text/plain": [
"┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
"┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
"┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
"│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m128\u001b[0m) │ \u001b[38;5;34m3,456\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_1 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m8,256\u001b[0m │\n",
"├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
"│ dense_2 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m65\u001b[0m │\n",
"└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Total params: 11,777 (46.00 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Total params: \u001b[0m\u001b[38;5;34m11,777\u001b[0m (46.00 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Trainable params: 11,777 (46.00 KB)\n",
"\n"
],
"text/plain": [
"\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m11,777\u001b[0m (46.00 KB)\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Non-trainable params: 0 (0.00 B)\n",
"\n"
],
"text/plain": [
"\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"model = build_model(input_dim=X_train_scaled.shape[1])\n",
"model.summary()"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Epoch 1/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m62s\u001b[0m 2ms/step - accuracy: 0.8986 - loss: 0.2663 - val_accuracy: 0.9128 - val_loss: 0.2386\n",
"Epoch 2/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m61s\u001b[0m 2ms/step - accuracy: 0.9132 - loss: 0.2369 - val_accuracy: 0.9150 - val_loss: 0.2329\n",
"Epoch 3/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m67s\u001b[0m 2ms/step - accuracy: 0.9167 - loss: 0.2298 - val_accuracy: 0.9177 - val_loss: 0.2269\n",
"Epoch 4/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m64s\u001b[0m 2ms/step - accuracy: 0.9186 - loss: 0.2263 - val_accuracy: 0.9194 - val_loss: 0.2247\n",
"Epoch 5/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m64s\u001b[0m 2ms/step - accuracy: 0.9199 - loss: 0.2236 - val_accuracy: 0.9203 - val_loss: 0.2235\n",
"Epoch 6/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m67s\u001b[0m 2ms/step - accuracy: 0.9207 - loss: 0.2222 - val_accuracy: 0.9224 - val_loss: 0.2197\n",
"Epoch 7/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m65s\u001b[0m 2ms/step - accuracy: 0.9215 - loss: 0.2203 - val_accuracy: 0.9209 - val_loss: 0.2216\n",
"Epoch 8/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m65s\u001b[0m 2ms/step - accuracy: 0.9218 - loss: 0.2199 - val_accuracy: 0.9228 - val_loss: 0.2176\n",
"Epoch 9/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m65s\u001b[0m 2ms/step - accuracy: 0.9224 - loss: 0.2185 - val_accuracy: 0.9221 - val_loss: 0.2191\n",
"Epoch 10/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m78s\u001b[0m 2ms/step - accuracy: 0.9229 - loss: 0.2178 - val_accuracy: 0.9230 - val_loss: 0.2180\n",
"Epoch 11/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m73s\u001b[0m 2ms/step - accuracy: 0.9231 - loss: 0.2173 - val_accuracy: 0.9230 - val_loss: 0.2170\n",
"Epoch 12/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m75s\u001b[0m 2ms/step - accuracy: 0.9234 - loss: 0.2169 - val_accuracy: 0.9240 - val_loss: 0.2163\n",
"Epoch 13/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m71s\u001b[0m 2ms/step - accuracy: 0.9236 - loss: 0.2163 - val_accuracy: 0.9238 - val_loss: 0.2162\n",
"Epoch 14/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m74s\u001b[0m 2ms/step - accuracy: 0.9239 - loss: 0.2158 - val_accuracy: 0.9244 - val_loss: 0.2150\n",
"Epoch 15/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m68s\u001b[0m 2ms/step - accuracy: 0.9240 - loss: 0.2154 - val_accuracy: 0.9240 - val_loss: 0.2161\n",
"Epoch 16/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m67s\u001b[0m 2ms/step - accuracy: 0.9240 - loss: 0.2156 - val_accuracy: 0.9234 - val_loss: 0.2171\n",
"Epoch 17/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m65s\u001b[0m 2ms/step - accuracy: 0.9244 - loss: 0.2146 - val_accuracy: 0.9249 - val_loss: 0.2151\n",
"Epoch 18/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m65s\u001b[0m 2ms/step - accuracy: 0.9243 - loss: 0.2148 - val_accuracy: 0.9236 - val_loss: 0.2163\n",
"Epoch 19/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m66s\u001b[0m 2ms/step - accuracy: 0.9247 - loss: 0.2137 - val_accuracy: 0.9247 - val_loss: 0.2144\n",
"Epoch 20/20\n",
"\u001b[1m34976/34976\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m70s\u001b[0m 2ms/step - accuracy: 0.9247 - loss: 0.2139 - val_accuracy: 0.9239 - val_loss: 0.2162\n"
]
}
],
"source": [
"history = model.fit(X_train_scaled, y_train,\n",
" epochs=20, batch_size=128,\n",
" validation_split=0.2)\n"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[1m43720/43720\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m46s\u001b[0m 1ms/step - accuracy: 0.9243 - loss: 0.2154\n",
"Validation Loss: 0.2155\n",
"Validation Accuracy: 0.9244\n"
]
}
],
"source": [
"# 모델 평가 (검증 데이터 기준)\n",
"loss, accuracy = model.evaluate(X_test_scaled, y_test)\n",
"\n",
"# 결과 출력\n",
"print(f\"Validation Loss: {loss:.4f}\")\n",
"print(f\"Validation Accuracy: {accuracy:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 22,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[1m43720/43720\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m28s\u001b[0m 636us/step\n"
]
}
],
"source": [
"y_test_pred = model.predict(X_test_scaled) # 테스트 데이터에 대해 예측\n",
"y_test_pred = (y_test_pred > 0.3549).astype(int) # 이진 분류이므로 0.5를 기준으로 0과 1로 변환\n"
]
},
{
"cell_type": "code",
"execution_count": 23,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"from sklearn.metrics import confusion_matrix\n",
"import seaborn as sns\n",
"import matplotlib.pyplot as plt\n",
"\n",
"# 혼동행렬 계산\n",
"cm = confusion_matrix(y_test, y_test_pred)\n",
"\n",
"# 시각화 (heatmap)\n",
"plt.figure(figsize=(6, 5))\n",
"sns.heatmap(cm, annot=True, fmt=\"d\", cmap=\"Blues\", xticklabels=[\"Normal\", \"Malicious\"], yticklabels=[\"Normal\", \"Malicious\"])\n",
"plt.ylabel('Actual')\n",
"plt.xlabel('Predicted')\n",
"plt.title('Confusion Matrix')\n",
"plt.show()\n"
]
},
{
"cell_type": "code",
"execution_count": 24,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\u001b[1m43720/43720\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m28s\u001b[0m 637us/step\n",
"최적의 임계값: 0.3844\n",
"Test Recall: 0.7656\n",
"Test Precision: 0.8790\n",
"AUC-PR: 0.8924\n",
"F1 Score : 0.8184\n",
"Accuracy: 0.9240\n",
"ROC-AUC: 0.9379\n"
]
}
],
"source": [
"from sklearn.metrics import precision_recall_curve, recall_score, precision_score, average_precision_score, f1_score, accuracy_score, roc_auc_score\n",
"\n",
"# 테스트 데이터에 대한 예측 확률 (양성 클래스)\n",
"y_test_pred_prob = model.predict(X_test_scaled)\n",
"\n",
"# 여러 임계값에서 F1 Score 계산\n",
"precisions, recalls, thresholds = precision_recall_curve(y_test, y_test_pred_prob)\n",
"f1_scores = 2 * (precisions * recalls) / (precisions + recalls + 1e-9) # 0으로 나누는 오류 방지\n",
"\n",
"# F1 Score가 최대가 되는 임계값 선택\n",
"best_threshold = thresholds[np.argmax(f1_scores)]\n",
"print(f\"최적의 임계값: {best_threshold:.4f}\")\n",
"\n",
"# f1을 활용한 임계값을 기준으로 클래스 예측\n",
"y_test_pred_class = (y_test_pred_prob >= best_threshold).astype(int)\n",
"\n",
"# Recall 계산\n",
"recall = recall_score(y_test, y_test_pred_class)\n",
"print(f\"Test Recall: {recall:.4f}\")\n",
"\n",
"# Precision 계산\n",
"precision = precision_score(y_test, y_test_pred_class)\n",
"print(f\"Test Precision: {precision:.4f}\")\n",
"\n",
"# AUC-PR 점수 계산\n",
"aucpr_score = average_precision_score(y_test, y_test_pred_prob)\n",
"print(f\"AUC-PR: {aucpr_score:.4f}\")\n",
"\n",
"# F1 Score 계산\n",
"print(f\"F1 Score : {f1_score(y_test, y_test_pred_class):.4f}\")\n",
"\n",
"# Accuracy 점수 계산\n",
"accuracy = accuracy_score(y_test, y_test_pred_class)\n",
"print(f\"Accuracy: {accuracy:.4f}\")\n",
"\n",
"# ROC-AUC 점수 계산\n",
"roc_auc = roc_auc_score(y_test, y_test_pred_prob)\n",
"print(f\"ROC-AUC: {roc_auc:.4f}\")\n"
]
},
{
"cell_type": "code",
"execution_count": 25,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"from sklearn.metrics import f1_score, accuracy_score\n",
"\n",
"# threshold 간단히 100개로 나눔\n",
"thresholds = np.linspace(0, 1, 100)\n",
"\n",
"f1_scores = []\n",
"accuracies = []\n",
"\n",
"for t in thresholds:\n",
" y_pred_class = (y_test_pred_prob >= t).astype(int)\n",
" f1_scores.append(f1_score(y_test, y_pred_class))\n",
" accuracies.append(accuracy_score(y_test, y_pred_class))\n",
"\n",
"best_idx = np.argmax(f1_scores)\n",
"best_threshold = thresholds[best_idx]\n",
"\n",
"# 그래프 그리기\n",
"plt.figure(figsize=(12, 6))\n",
"plt.plot(thresholds, f1_scores, label='F1 Score', color='blue')\n",
"plt.plot(thresholds, accuracies, label='Accuracy', color='green')\n",
"plt.axvline(x=best_threshold, color='red', linestyle='--', label=f'Best Threshold = {best_threshold:.2f}')\n",
"\n",
"plt.xlabel('Threshold')\n",
"plt.ylabel('Score')\n",
"plt.title('F1 Score & Accuracy vs Threshold')\n",
"plt.legend()\n",
"plt.grid(True)\n",
"plt.tight_layout()\n",
"plt.show()\n"
]
},
{
"cell_type": "code",
"execution_count": 26,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"PermutationExplainer explainer: 1001it [01:10, 13.74it/s] \n"
]
},
{
"data": {
"image/png": "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",
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import shap\n",
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"\n",
"# DataFrame 컬럼명 지정 (중요!)\n",
"X_test_scaled_df = pd.DataFrame(X_test_scaled, columns=X_test.columns)\n",
"\n",
"# 샘플 추출\n",
"X_sample = X_test_scaled_df.sample(n=1000, random_state=42)\n",
"\n",
"# SHAP 값 계산\n",
"explainer = shap.Explainer(model, X_sample)\n",
"shap_values = explainer(X_sample)\n",
"\n",
"# 평균 절대 SHAP 값 계산\n",
"mean_abs_shap = np.abs(shap_values.values).mean(axis=0)\n",
"\n",
"# DataFrame으로 정리 (feature name + 중요도)\n",
"shap_df = pd.DataFrame({\n",
" 'feature': X_sample.columns,\n",
" 'importance': mean_abs_shap\n",
"})\n",
"\n",
"# 중요도 기준 정렬\n",
"shap_df = shap_df.sort_values(by='importance', ascending=False)\n",
"\n",
"# 상위 N개 선택\n",
"top_n = 30\n",
"shap_df_top = shap_df.head(top_n)\n",
"\n",
"# Bar plot 그리기\n",
"plt.figure(figsize=(10, 8))\n",
"bars = plt.barh(shap_df_top['feature'][::-1], shap_df_top['importance'][::-1], color='skyblue')\n",
"\n",
"# 수치(label) 추가\n",
"for bar in bars:\n",
" width = bar.get_width()\n",
" plt.text(width + 0.001, bar.get_y() + bar.get_height()/2, f'{width:.4f}', va='center')\n",
"\n",
"plt.xlabel('Mean |SHAP value|')\n",
"plt.title('Top SHAP Feature Importances')\n",
"plt.tight_layout()\n",
"plt.show()\n"
]
},
{
"cell_type": "code",
"execution_count": 27,
"metadata": {},
"outputs": [],
"source": [
"# import tensorflow as tf\n",
"# from sklearn.metrics import roc_auc_score\n",
"\n",
"# # 중요도 기준으로 특성 제거\n",
"# # numpy 배열을 DataFrame으로 변환 (컬럼명 유지)\n",
"# X_train_scaled_df = pd.DataFrame(X_train_scaled, columns=X_train.columns)\n",
"# X_test_scaled_df = pd.DataFrame(X_test_scaled, columns=X_test.columns)\n",
"\n",
"# # 중요도 낮은 특성 제거\n",
"# threshold = 0.0001\n",
"# low_importance_features = shap_df[shap_df['importance'] < threshold]['feature'].tolist()\n",
"\n",
"# print(f\"제거할 특성 수: {len(low_importance_features)}\")\n",
"# print(f\"제거 된 특성 : {low_importance_features}\")\n",
"\n",
"# # 제거한 특성 반영\n",
"# X_train_reduced = X_train_scaled_df.drop(columns=low_importance_features)\n",
"# X_test_reduced = X_test_scaled_df.drop(columns=low_importance_features)\n",
"\n",
"# # numpy로 변환\n",
"# X_train_np = X_train_reduced.to_numpy()\n",
"# X_test_np = X_test_reduced.to_numpy()\n",
"\n",
"\n",
"# model = build_model(input_dim=X_train_np.shape[1])\n",
"\n",
"# history = model.fit(X_train_np, y_train,\n",
"# epochs=30, batch_size=256,\n",
"# validation_split=0.2)\n",
"\n",
"\n"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {},
"outputs": [],
"source": [
"# # 예측 및 평가\n",
"# y_pred_reduced = model.predict(X_test_np).ravel()\n",
"# roc_reduced = roc_auc_score(y_test, y_pred_reduced)\n",
"\n",
"# print(f\"SHAP 기반 특성 제거 후 ROC AUC: {roc_reduced:.4f}\")"
]
},
{
"cell_type": "code",
"execution_count": 29,
"metadata": {},
"outputs": [],
"source": [
"# model.save(\"White_list_model.keras\")"
]
},
{
"cell_type": "code",
"execution_count": 30,
"metadata": {},
"outputs": [],
"source": [
"# from tensorflow.keras.models import load_model\n",
"\n",
"# model = load_model('White_list_model.keras')"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {},
"outputs": [],
"source": [
"# url = \"https://www.11st.co.kr/?gad_source=1&gclid=CjwKCAjwn6LABhBSEiwAsNJrjqtgvZC-SH5p6dnrQwDUjxMsWp96y_Jouf39QXALiirxR5KErWDNzhoCHZgQAvD_BwE&gbraid=0AAAAADQ_R8LZJlCP90c4HNMof-oD17G2u&utm_term=11+%B9%F8%B0%A1&utm_campaign=%B1%B8%B1%DBPC_11%B9%F8%B0%A1&utm_source=%B1%B8%B1%DB_PC_S&utm_medium=%B0%CB%BB%F6\"\n",
"\n",
"# features = extract_features(url)\n",
"\n",
"# input_df = pd.DataFrame([features])\n",
"\n",
"# scaler = MinMaxScaler()\n",
"# scaler.fit(input_df)\n",
"\n",
"# scaled_input = scaler.transform(input_df)"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {},
"outputs": [],
"source": [
"# prediction = model.predict(scaled_input)\n",
"\n",
"# # 예측값 출력 (확률 값)\n",
"# predicted_prob = prediction[0][0]\n",
"\n",
"# # 임계값 기준으로 라벨 결정\n",
"# predicted_label = int(predicted_prob > 0.4174)\n",
"\n",
"# # 결과 출력\n",
"# print(f\"예측 확률: {predicted_prob:.4f}\")\n",
"# print(f\"예측 결과: {'악성' if predicted_label else '정상'}\")\n"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": ".venv",
"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": "3.11.9"
}
},
"nbformat": 4,
"nbformat_minor": 2
}