on 06-14-2017 4:11 AM
I'm new to SAP Predictive Analytics.
Is there any way to use SAP Predictive Analytics time series analysis modeler to rolling predict each data point?
for example, I have a data set with n=500 and p=10,and I want to predict index 401~500, when I predict index 401, I use index 1~400 as training data set to training model, when I predict index 402, I use index 1~401 as training data set to training model,...,when I predict index 500, I use index 1~499 as training data set to training model.
If anyone knows how to did this automatically, please don't hesitate to help me, thanks.
Hi, we do not currently have trial version of the Predictive Factory. You have to be licensed for it. You can check our webinars & tutorials. See links on our community page https://www.sap.com/community/topic/predictive-analytics.html. KInd regards Antoine
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hi, please avoid cross-posting. Post your questions only once as per SAP community guidelines. Kind regards Antoine
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Can you please define n and p in your example? I understand n means rows and p means variables. Is that correct?
I understand you receive a new event every minute, is that correct?
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
Hello, what is the typical time interval associated to each or your data point? Using Predictive Factory, you can set up a task that will refresh the forecasts based on the last historical data points you just received in your database. It means I can re-learn the forecast & apply the model based on my last historical 500 points. Kind regards, Antoine
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
User | Count |
---|---|
90 | |
10 | |
10 | |
10 | |
7 | |
7 | |
6 | |
5 | |
4 | |
3 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.