Chapter 6: Humidity Forecasting with Classical Methods
In this chapter, we will build the first forecasting model ready for deployment, looking at how data can be recorded from sensors attached to an Arduino controller.
You’ll learn how KNIME Analytics Platform, as well as KNIME Server, can connect to sensors via REST endpoints, the basics of time series data cleaning and pre-processing, the different options for data granularity levels, and several classic and easy-to-use models that can be used for forecasting. Finally, we’ll look at some simple ways the model can be deployed for real-world applications with KNIME, such as writing model predictions to databases or saving trained models and workflows for later use.
We will cover the following main topics in the chapter:
- The importance of predicting the weather
- Streaming humidity data from an Arduino sensor
- Resampling and granularity
- Training and deployment
By the end of this chapter,...
