ml_model.h5

import numpy as np
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import CountVectorizer

# Dataset contoh (payload dan label untuk serangan SQL Injection)
payloads = [
    "' OR 1=1 --", 
    "' UNION SELECT null, username, password FROM users --", 
    "<script>alert('XSS')</script>", 
    "1' OR '1'='1' --", 
    "' DROP TABLE users;"
]
labels = [1, 1, 0, 1, 1]  # 1 = serangan berbahaya, 0 = tidak berbahaya

# Mengubah payload menjadi fitur numerik
vectorizer = CountVectorizer()
X = vectorizer.fit_transform(payloads).toarray()

# Membagi data menjadi pelatihan dan pengujian
X_train, X_test, y_train, y_test = train_test_split(X, labels, test_size=0.3, random_state=42)

# Membangun model jaringan saraf
model = Sequential()
model.add(Dense(128, input_dim=X_train.shape[1], activation='relu'))
model.add(Dense(64, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

# Menyusun model
model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])

# Melatih model
model.fit(X_train, y_train, epochs=10, batch_size=2, validation_data=(X_test, y_test))

# Menyimpan model ke file H5
model.save('ml_model.h5')

print("Model telah disimpan dalam 'ml_model.h5'")