Chapter 11: Testing Data for Time Series Compliance
Predicting values with a time series requires that we have historical data to analyze whether past values have a relationship with present ones and whether this relationship can be useful to predict future values. To validate this, we have to test the autocorrelation of the data. This chapter's purpose is to train professionals responsible for doing purchase planning and inventories, as well as those who need to do forecasts. To do effective planning based on data, we first have to validate that we have a good fit to do this kind of analysis. We need data to contain past values that correlate with present ones to predict future ones. We can create charts to see whether there is predictable behavior in our data and then run a statistical test to confirm that. To do this research, we will be doing the following in this chapter:
- Visualizing seasonal trends
- Researching autocorrelation – past values' influence...