I'm experimenting with data mining, trying to use the Time Series algorithm to predict future revenue based on existing data.
Following the example Brian Knight showed in his recent webinar on Data Mining, I created a view of my data like the vTimeSeries view, that is, aggregated by a few columns. However, when I select the multiple columns that define the aggregation for the keys in the data mining model and attempt to process the model, I keep getting an error "there are multiple key columns defined".
I really need the multiple keys, they define the aggregation. And, Brian selected two keys (region and time) in the example, but didn't seem to encounter this error. I haven't found any explanation for this on the web, so I thought I would post here and see if anyone had an explanation for this, and a solution.
You can have only one time key column.
Check this from http://msdn.microsoft.com/en-us/library/ms174923.aspx
The requirements for a time series model are as follows:
A single key time column Each model must contain one numeric or date column that is used as the case series, which defines the time slices that the model will use. The data type for the key time column can be either a datetime data type or a numeric data type. However, the column must contain continuous values, and the values must be unique for each series. The case series for a time series model cannot be stored in two columns, such as a Year column and a Month column.
Yeah, I can see why that must be, but I was a bit stubborn about it, given that I saw Brian select two key columns in his example. The other column (Region) must have been a "Key Series" column in his example, I surmise after reading the link you posted. That link is quite helpful, thanks for the reference.