How good is Splunk in terms of forecasting and comfort to separate Train Dataset and Test data set from same source data. Do we also get r2 and p value information to validate model under consideration
The Splunk Machine Learning Toolkit has a forecasting assistant with all the features you're asking about: withholding, validation statistics, and operationalization options.
Thanks mate.
Is there any way to also investigate p values
In the context of forecasting, where you have only one time series, are making only one forecast, and are not making assumptions about the distribution of the underlying data, p-values have dubious value. We decided not to build it into the assistant.