The importance of predicting the weather
Forecasting the weather is something people have been doing for nearly forever, with a huge amount of different approaches. For this chapter, we use humidity as an accessible and relatable example. The most sophisticated modern methods for forecasting the weather, temperature, and humidity use complicated physical simulations alongside data science techniques, and we can’t really hope to outperform them for long-term forecasts.
Despite this, we will see that with a concisely defined purpose, even the simplest models can prove useful. In this chapter, we will focus on generating forecasts only a few units (in our case, hours) into the future. If the humidity is in the process of spiking, we will want to bring our sensors, and whatever they are attached to, inside before it rains. In later chapters, we will extend the forecast horizon to predict days or weeks into the future.
Building Concise Data Science Projects
When building...
