| import tensorflow as tf | |
| import pandas as pd | |
| class SecIDSModel: | |
| def __init__(self, model_path="SecIDS-CNN.h5"): | |
| # Load the trained model | |
| self.model = tf.keras.models.load_model(model_path) | |
| def predict(self, data): | |
| # Preprocess data if needed (assume data is a Pandas DataFrame) | |
| processed_data = self.preprocess_data(data) | |
| # Make predictions | |
| predictions = self.model.predict(processed_data) | |
| # Convert predictions to readable format if needed | |
| return ["Attack" if pred > 0.5 else "Benign" for pred in predictions] | |
| def preprocess_data(self, data): | |
| # Placeholder for preprocessing logic, adjust according to your needs | |
| # For example, you may need to scale or reshape data | |
| return data.values | |