Linear Regression is a supervised statistical learning method that models the relationship between one or more input features and a continuous output variable by fitting a linear equation that minimizes prediction error, commonly applied in sensing research to tasks such as estimating crowd counts or vehicle counts from signal-derived features. In the context of WiFi CSI and Bluetooth-based sensing, it serves as a foundational baseline or lightweight inference model that maps extracted environmental or signal features directly to scalar quantities of interest, making it particularly valuable for validating more complex approaches and benchmarking performance in regression tasks. Key variants relevant to this domain include simple linear regression with a single predictor, multiple linear regression incorporating several CSI or Bluetooth features simultaneously, and regularized forms such as Ridge and Lasso regression, which help mitigate overfitting when feature dimensionality is high relative to available training samples.
Source Papers
- Bluetooth-Based Vehicle Counting: Bridging the Gap to Ground-Truth With Machine Learning ↗ — Bluetooth-Based Vehicle Counting: Bridging the Gap to Ground
- CrossSense: Towards Cross-Site and Large-Scale WiFi Sensing ↗ — CrossSense: Towards Cross-Site and Large-Scale WiFi Sensing
- Occupancy Prediction in IoT-Enabled Smart Buildings: Technologies, Methods, and Future Directions ↗ — Occupancy Prediction in IoT-Enabled Smart Buildings: Technol
- Wi-CaL: WiFi Sensing and Machine Learning Based Device-Free Crowd Counting and Localization ↗ — Wi-CaL: WiFi Sensing and Machine Learning Based Device-Free