Creating a forecast by fitting a curve to historic data
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Mute
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Notify Moderator
Hello everyone, i would love to learn how to use the predictive analytics tools in alteryx (perhaps a neural network?) but so far i havent figured out how to apply those tools to my data.
So instead, i am trying to manually fit a curve to my historic data, and use it as a model for future projections.
So far my historic data appears to fit a geometric pattern ( A * B^(n-1) )
Now what i need to do, is figure out how to fit that curve (green line) systematically onto the blue line.. does anyone know how i can get alteyx to find the best values for A and B to best fit the blue line?.. Once i have that i believe i can then cross multiply those values to fit it onto the orange line.. That would become my "forecast"
Right now i've manually set A = 1.3, and B = 1.02.. but i want alteryx to tell me those values, rather than manually plugging them into the formula
attached is a sample set of data, and a section of the workflow
Solved! Go to Solution.
- Labels:
- Predictive Analysis
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Notify Moderator
This isn't the solution I want, but I've built a rudimentary "forecast" by just multiplying the historic data by the ratio of actual data divided by historic data..
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Notify Moderator
![](/skins/images/0052A40480681DBBC707042CBFDD66A8/responsive_peak/images/icon_anonymous_message.png)