In my earlier article, we explored the fundamentals of time collection forecasting with sktime, find out how to leverage this highly effective library for simple forecasting duties. Now, it’s time to take our journey additional and dive into the superior strategies that may enable you to optimize your forecasts and enhance their accuracy. On this follow-up, we’ll discover find out how to construct extra refined fashions, tune hyperparameters, and even do mannequin structure search with sktime.
First, for a straightforward begin, let me exhibit the essential sktime workflow once more. This time, we are going to use the Longley dataset, which is a part of sktime (BSD-3 license). It incorporates varied US macroeconomic variables from the years 1947 to 1962 and appears like this: