I'm just getting started with Eureqa, but even running a session against some predictable demo data is already frustrating.
For example, I have 100 products with sales data for the first 30 days. I would like to see if there is a correlation in the first 5 days of sales what the sales will look like on day 30.
However, even when I run predictable samples like every day is (prior day + 10) units, I get results back like "day30 = day5". I've played with both cumulative figures (d1: 10, d2: 20, d3: 30) and daily values (d1: 10, d2: 11, d3: 12) and neither really find a valuable solution.
Am I completely missing something? I'd just like to understand this better before I throw a big data set at it.
[link] [1 comment]