Forecasting Internal Temperature in a Home with a Sensor Network

We forecast internal temperature in a home with sensors, modeled as a linear function of recent sensor values. The Smart* Project provides publicly available data from an inhabited home over a three month period, reporting on 38 sensors including environmental readings, circuit loads, motion detectors, and switches controlling lights and fans. We select 13 of these sensors that have some influence on the internal temperature, and create forecasts that are accurate to within about 1.6°F (0.9°C) over the next six hours. Temperature prediction is important for saving energy while maintaining comfortable conditions in the home