Scribd Logo
Upload

Welcome to Scribd!

  • 167 views

    SSAS - Data Mining

    Uploaded by

    jeevan619
    This document describes a case study using SQL Server Analysis Services 2008 for data mining. It outlines the steps to create several different types of mining models using the AdventureWorksDW sample database: 1. A decision tree model to predict whether customers will purchase bikes based on demographic data in the TargetedMailing structure. 2. A clustering and naive bayes model in the modified MiningStructure. 3. Using the decision tree model to generate predictions by querying the ProspectiveBuyer table. 4. A time series forecasting model in the Forecasting structure to predict future Quantity values based on historical TimeIndex and ModelRegion data in the vTimeSeries table.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as DOC, PDF, TXT or read online from Scribd

    SSAS - Data Mining

    Uploaded by

    jeevan619
    167 views4 pages
    This document describes a case study using SQL Server Analysis Services 2008 for data mining. It outlines the steps to create several different types of mining models using the AdventureWorksDW sample database: 1. A decision tree model to predict whether customers will purchase bikes based on demographic data in the TargetedMailing structure. 2. A clustering and naive bayes model in the modified MiningStructure. 3. Using the decision tree model to generate predictions by querying the ProspectiveBuyer table. 4. A time series forecasting model in the Forecasting structure to predict future Quantity values based on historical TimeIndex and ModelRegion data in the vTimeSeries table.

    Original Title

    SSAS - Data Mining[1]

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    DOC, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document describes a case study using SQL Server Analysis Services 2008 for data mining. It outlines the steps to create several different types of mining models using the AdventureWorksDW sample database: 1. A decision tree model to predict whether customers will purchase bikes based on demographic data in the TargetedMailing structure. 2. A clustering and naive bayes model in the modified MiningStructure. 3. Using the decision tree model to generate predictions by querying the ProspectiveBuyer table. 4. A time series forecasting model in the Forecasting structure to predict future Quantity values based on historical TimeIndex and ModelRegion data in the vTimeSeries table.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    167 views4 pages

    SSAS - Data Mining

    Uploaded by

    jeevan619
    This document describes a case study using SQL Server Analysis Services 2008 for data mining. It outlines the steps to create several different types of mining models using the AdventureWorksDW sample database: 1. A decision tree model to predict whether customers will purchase bikes based on demographic data in the TargetedMailing structure. 2. A clustering and naive bayes model in the modified MiningStructure. 3. Using the decision tree model to generate predictions by querying the ProspectiveBuyer table. 4. A time series forecasting model in the Forecasting structure to predict future Quantity values based on historical TimeIndex and ModelRegion data in the vTimeSeries table.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    of 4

    SSAS 2008 Data Mining

    Case Study

    1. Getting Started
    Create a new Analysis Services Project Create a Data Source based on AdventureWorks DW Databast Create a Data Source View with the following tables:

    dbo.ProspectiveBuyer dbo.vAssocSeqLineItems dbo.vAssocSeqOrders dbo.vTargetMail dbo.vTimeSeries Set up a many-many relationship between .vAssocSeqOrders and vAssocSeqLineItems tables (Select the OrderNumber column in the vAssocSeqLineItems table.Drag the column to the vAssocSeqOrders table, and put it on the OrderNumber column.)

    2. Creating a Targeted Mailing Mining Model Structure


    Using Data mining Wizard, set up as follows: Method: From existing relational DB or DW Data mining Technique: Microsoft Decision Trees Data Source: Adventure Works DW Table Types:vTargetMail (Case column) Training Data : verify that the check box in the Key column is selected next to the CustomerKey column. If the source table from the data source view indicates a key, the Data Mining Wizard automatically chooses that column as a key for the model Bike Buyer Column Input, Predictable Select the Input check boxes next to the following columns: Age CommuteDistance EnglishEducation EnglishOccupation FirstName Gender GeographyKey

    Pg : 1

    SSAS 2008 Data Mining


    Case Study

    HouseOwnerFlag LastName MaritalStatus NumberCarsOwned NumberChildrenAtHome Region TotalChildren YearlyIncome On the Specify Columns' Content and Data Type page, click Detect Specify names for Mining Structure and Mining model name Ensure Allow Drill Through is selected

    3. Modifying the Mining Structure Creating a clustering model


    Create two new modes as: Algorithm:Microsoft Clustering Algorithm: Microsoft Naive Bayes

    For the Microsoft Naive Bayes, set BikeBuery column as Input and Predictable Build the Project and Deploy Process the Data Mining Structure

    4. Creating Predictions
    On the Mining Model Prediction tab of Data Mining Designer, in the Mining Model box, click Select Model. Select the following: Select Model: Select the decision tree model

    Select Input Tables->Select Case Table->Table/View Name: ProspectiveBuyer To build the Predicition Query, Set the following In the Source column in the grid on the Mining Model Prediction tab Pg : 2

    SSAS 2008 Data Mining


    Case Study

    PropspetiveBuyer / ProspectAlternateKey Your Decision Tree/ Bike Buyer column PredictionFunction/PredictProbability [<Your Decision Tree>.[Bike Buyer] (Criteria/Argeument Column:

    You can run the query by clicking the arrow next to the Switch to query design view / Switch to query result view button, the first button on the toolbar, and then selecting Result.

    5. Creating a Forecasting Model


    Using the Data Mining Wizard, set the following: Definition warehouse Method: From existing relational database or data

    Data Mining Technique: Microsoft Time Series Data Source View: Adventure Works DW TableTypes: select the Case check box next to the vTimeSeries table

    TrainingData: select the Key check boxes next to the TimeIndex and ModelRegion columns, Select the Input and Predictable check boxes next to the Quantity column. Specify Columns' Content and Data Type: The Time Index column is automatically designated as a key time column and the Model Region column is designated as a key column. This means that a separate time series model will be built for each unique entry in the Model Region column. The values in the Time Index column must be unique across individual values in the Model Region column. Mining structure name: Forecasting Mining model name: Forecasting To modify the model:

    On the Mining Structure tab of Data Mining Designer, in the Data Source View pane, select the Amount column in the vTimeSeries table. Drag the Amount column from the Data Source View pane into the list of columns for the Forecasting structure. In the Forecasting column of the grid on the Mining Models tab, click the cell that corresponds to the Amount row. Select Predict. The Amount column is now both an input column and a predictable column.

    Pg : 3

    SSAS 2008 Data Mining


    Case Study

    In the Mining Models tab, right-click Forecasting, and select Set Algorithm Parameters. The Algorithm Parameters dialog box opens. o In the Value column, set PERIODICITY_HINT to {12}, and then click OK. Process the model

    *** End of Document ***

    Pg : 4

    You might also like