Codelord01/sensor_binary

1---2license: apache-2.03language: en4library_name: keras5tags:6- intrusion-detection7- cyber-physical-systems8- iot-security9- lstm10- time-series11- cybersecurity12datasets:13- ToN_IoT14---15 16# ClimIDS: Sensor-Layer Intrusion Detection System17 18This model card is for **ClimIDS**, a lightweight, LSTM-based intrusion detection system (IDS) for the physical sensor layer of IoT deployments.19 20## Model Description21ClimIDS analyzes time-series data from environmental sensors (temperature, pressure, humidity) to detect anomalies in climate-monitoring systems. Its lightweight architecture (~5,000 parameters) makes it suitable for edge devices.22 23- **Architecture:** `LSTM -> Dropout -> Dense -> Dense (Sigmoid)`24- **Dataset:** Trained on `IoT_Weather` subset of ToN_IoT25- **Performance:** 98.81% accuracy, 99.7% attack recall26 27## Intended Use28- **Primary Use:** Real-time binary classification of sensor telemetry29- **Input:** `(batch_size, 10, 3)` — features `[temperature, pressure, humidity]`, normalized30- **Output:** Float between 0.0 (Normal) and 1.0 (Attack), threshold 0.531 32## How to Use33```python34import tensorflow as tf35import numpy as np36from huggingface_hub import hf_hub_download37 38MODEL_PATH = hf_hub_download("Codelord01/sensor_binary", "sensor_binary.keras")39model = tf.keras.models.load_model(MODEL_PATH)40model.summary()41 42sample_data = np.random.rand(1, 10, 3).astype(np.float32)43prediction_prob = model.predict(sample_data)44predicted_class = 1 if prediction_prob > 0.5 else 045print(f"Prediction Probability: {prediction_prob:.4f}")46print("Anomaly Detected" if predicted_class == 1 else "Normal Conditions")47