You can deploy this model and start earning money today!

Codelord01/
binary_model

Name: binary_model
Rating: 2.5 (14 reviews)

Views

No views yet

Binary Network-Layer Cyber-Physical IDS

A hybrid CNN-BiLSTM model for real-time binary network intrusion detection in IoT environments.
This model acts as the first line of defense by quickly distinguishing between malicious and legitimate traffic.

Model Description

Architecture: Conv1D -> ... -> Bidirectional LSTM -> Dense -> Dense (Sigmoid)
Dataset: Balanced subset of CICIoT2023
Performance: 99.9997% accuracy
Limitations: Validated only on CICIoT2023-like network traffic; may not detect novel attack types. Input must be normalized.
Training Information:
- Optimizer: Adam
- Loss: Binary Cross-Entropy
- Balanced dataset: 2 million samples (1M benign, 1M attack)

Intended Use

Primary Use: Real-time network intrusion detection
Input: (batch_size, 10, 46) — 46 network flow features, normalized
Output: Float between 0.0 (Benign) and 1.0 (Attack), threshold 0.5

How to Use

python

import tensorflow as tf
import numpy as np
from huggingface_hub import hf_hub_download

# Download the model from Hugging Face
MODEL_PATH = hf_hub_download("Codelord01/binary_model", "binary_model.keras")
model = tf.keras.models.load_model(MODEL_PATH)
model.summary()

# Prepare a sample input: 1 sample, 10 timesteps, 46 features
sample_data = np.random.rand(1, 10, 46).astype(np.float32)

# Make a prediction
prediction_prob = model.predict(sample_data)
predicted_class = 1 if prediction_prob > 0.5 else 0

print(f"Prediction Probability: {prediction_prob:.4f}")
print("Malicious Traffic Detected" if predicted_class == 1 else "Benign Traffic")


@mastersthesis{ababio2025multilayered,
  title={A Multi-Layered Hybrid Deep Learning Framework for Cyber-Physical Intrusion Detection in Climate-Monitoring IoT Systems},
  author={Awuni David Ababio},
  year={2025},
  school={Kwame Nkrumah University of Science and Technology}
}