Prioritize...
Once you have completed this section, you should be able to define a categorical weather-dependent problem.
Read...
Categorical variables take on one of a limited (at least two) choice of fixed possible outcomes. In weather, categorical variables include hurricane categories, tornado strength, and precipitation type. You might think that categorical forecasts are not as important as say a probability forecast, but as we saw in the previous lessons, many probabilistic forecasts can become categorical forecasts, particularly when used to make a cost-based decision. For the same reason, many times a real-world scenario will require a categorical answer, not the value of a continuous variable. Read on to learn about what problems require categorical forecasts.
Overview
For a categorical forecast, there will be two or more possible outcomes. The goal is to predict the outcome category from the weather data. For example, one question may be whether a river will rise enough to overtop a levee. There are two possible outcomes: the river will overtop the levee, or it won’t. Flood preparation decisions would depend on the odds of the levee being overtopped.
Another example of a categorical weather-dependent forecast is with respect to aviation. Based on certain weather conditions, pilots need to know what set of flight rules apply. The rules are Visual Flight Rules (VFR), Marginal VFR (MVFR), Instrument Flight Rules (IFR), and Low Instrument Flight Rules (LIFR). You can learn more about these rules here. Since there are four different flight rules, we have four possible outcomes. Fuel reserve decisions for an executive jet would depend on the odds of each of these categories occurring.