SmartHome-AI-IoT Regressor Model

Overview

The SmartHome-AI-IoT Regressor Model predicts the desired air-conditioner temperature setting based on smart-home contextual information.

The model enables adaptive HVAC control by considering environmental and user-related conditions.

Authors

Belnadino Mgimba
Department of Computer Science, Durham University, Durham, United Kingdom

Anish Jindal
Department of Computer Science, Durham University, Durham, United Kingdom

Project

This model is part of the:

SmartHome-AI-IoT Framework
"A Natural Language Understanding, Low-Latency, Context-Aware, Multi-Device Control and Self-Learning SmartHome IoT Framework"

Developed for research on:

  • Natural language understanding for IoT
  • Context-aware smart-home automation
  • Edge AI deployment
  • Human-in-the-loop adaptive systems

Intended Use

Applications include:

  • Smart HVAC automation
  • Context-aware temperature control
  • Energy-efficient smart homes
  • Edge IoT systems

Model Architecture

Model: XGBoost Regression

Input features:

  • Room location
  • Time period
  • Ambient temperature
  • Occupancy

Output: Predicted AC temperature

Example:

Input: Room: Bedroom Time: Night Ambient temperature: 22°C Occupancy: 1

Output: AC temperature: 23°C

Dataset

Training data generated from:

SmartHomeIoTNLU Dataset

Dataset characteristics:

  • 200,320 smart-home events
  • Environmental context
  • Device states
  • AC temperature parameters

Training Configuration

XGBoost parameters:

Parameter Value
Trees 300
Maximum depth 6
Learning rate 0.05
Objective Squared Error

Performance

Metric Value
MAE 1.2349
RMSE 1.7725

Deployment

The model can be deployed in:

  • Raspberry Pi edge gateway
  • Smart-home controllers
  • IoT automation frameworks

Limitations

  • Temperature preferences are generated from synthetic rules.
  • Real occupants may have different comfort preferences.
  • Multi-user preference conflicts are not considered.

Citation

SmartHome-AI-IoT Framework

License

apache-2.0

Downloads last month
11
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support