data mining using SSAS design question

  • Hello -

    I have a design question and want to understand whether my problem could be solved using SSAS data mining and if yes, which algorithm will be useful.

    I have a revenue fact table with dimension keys for say, dim A, dim B, dim C and dim D along with date dim. ( I have already built a cube for this) I want to predict my revenue for dim A and also simultaneously see effect of dim B, dim C and dim D on the behavior of this prediction. so, for example I want to see prediction of my revenue by region. But then I should be able to select a product and see the prediction by region for that product and so on for other dimensions.

    I tried time series algorithm but found that i can have one time series with one date dim and one other dimension only. I can't have dim A and dim B simultaneously on time series along with dim date. (Is it true? or I'm missing something).

    What other possible algorithms i can use for this kind of scenario and how to model that in SSAS?

    Any help is appreciated.



  • would it be possible to combine the 4 dimensions in the DSV and then create 1 "super" cube-dimension for use in the dm model?

  • As long as Dim A-D are all linked together through the fact table, you should be able to add Dimension A as the initial Dimension and the rest of the Dimensions as nested Dimensions.

  • One key aspect is there should be that cannot be handled by a single computer. Other aspects that are proposed to defined big data is that the data is heterogeneous that it is evolving , and that it is decentralized and distributed.

  • your on investment. Salford Systems' Rapid Response Data Mining[/url] center plans and executes demonstration projects and proof-of-concept studies in as little as four weeks. Value assessments of large scale data mining projects are generated quickly and cost-effectively on our six high speed servers with massive storage capacity.

  • A database is a collection of records or data that is stored in a computer system. The structure of a database is dependent on how the data is organized, according to a particular database model. Today we commonly use a relational database model. Other models include a hierarchical model and the network model..

  • The progress made in hardware technology allows today’s

    computer systems to store very large amounts of data.:-D

  • When refreshing data that is persisted, the OnPropertyChanged event is raised for every property in every data object, this can cause a lot of slowdown.

  • The data mining role also extends into engineering, in particular infrastructure and logging. A few examples that have been developed .

Viewing 9 posts - 1 through 8 (of 8 total)

You must be logged in to reply to this topic. Login to reply