Let’s say you want to forecast future revenue. You could fill in a month-to-month or week-to-week spreadsheet template. However, these are very subjective methods of trying to predict the future and often wildly inaccurate.
A better method would be to create predictive models using time series analysis. Time series predictive models are a specific subset of statistical models that deal with data that are ordered by and dependent on time. On a graph, the x-axis would be time, with the y-axis the thing (revenue, in this case) you are measuring.
In this post, I will give an overview of one type of time series models and describe how they work in simple language. I will describe how to implement the models in BigQuery ML.