Logistic Regression is a supervised statistical classification algorithm that models the probability of a discrete class label by applying a logistic (sigmoid) function to a linear combination of input features, making it well-suited for binary or multiclass categorization tasks such as distinguishing activity types, occupancy states, or spatial sections from CSI amplitude or phase features. In WiFi sensing research, it serves as a lightweight, interpretable baseline classifier that requires minimal computational resources, enabling deployment on constrained hardware like ESP32 modules and providing a reproducible performance benchmark against which more complex models such as neural networks can be fairly compared. Key variants include One-vs-Rest (OvR) and multinomial logistic regression for handling multiclass problems, as well as regularized forms (L1/Lasso and L2/Ridge penalties) that mitigate overfitting when the number of CSI subcarrier features is large relative to the number of labeled training samples.
Source Papers
- A survey on CSI-based Wi-Fi sensing datasets and models with a focus on reproducibility ↗ — A survey on CSI-based Wi-Fi sensing datasets and models with
- Human Activity Recognition via Wi-Fi and Inertial Sensors With Machine Learning ↗ — Human Activity Recognition via Wi-Fi and Inertial Sensors Wi
- 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