Dear fellow analysts,
My question relates to the modes of model evaluation available with
SSAS. For example, Logistic regression in SSAS is evaluated using Lift
Chart. This is good to compare visually. But is there a way to get hard

I have a problem with this variable importance. Normally we are defining minimum suport and the confidence so what is this variable called importance? The book says it is to measure the quality of a rule. Are we taking quality rule or frequent rule. Pls help me to clarify on this ...

With the association rule the normal procedure is to defining a minimum support and the minimum confidence. But with Microsoft association rule algorithm there is a variable called importance. And the book says it is used to measure the quality and the usefulness of the rule. ...

I'm trying to set the criteria for a datetime field for all dates not older
than 60 days from the current system date.
GetDate() -60 doesn't work.
How do I do this?

when i clike the "process" of a newly-built mining structure , it pops-out an alarm which contains "The server content appears to be out of date,would you like to build and deploy the project first "
and when i click "run" after it was deployed successfully , it failed everytime , ...

Dear fellows,
 I have a Windows Server 2003 with SQL Server 2000 (Standard Edition)
installed and the OLAP Service(MSSQLServerOLAPServices) is on. and I
can do MDX Query on the "MDX Sample Application" on the server.

Is there a way to force decision trees algorithm to split on a selected
column instead of the predicted default?
If no, any workarounds?
CheSs

In Data Mining Tutorial,there is an example using EM clustering algorithm.
I know EM algorithm models continuous attributes with Gaussian models.But how Clustering Algorithm deal with these discrete attributes?
thanks.

I try the following MDX query before I apply to my application. I use the Microsoft SQL Server Managment Studio. I have results with this DMX
DMX query:
Select Flattened (SELECT [SUPPORT], ATTRIBUTE_VALUE FROM NODE_DISTRIBUTION), NODE_DESCRIPTION from all_under_tree_mining.Content ...

Has anyone ever tried to return the y-intercept for the equation y = mx + b?
In this case, y is fldSquareFeet and x is fldPrice.  Can this be done in a
View?  If so, could you please provide some sample SQL?
Thanks!!

I'm using SQL Server 2005 AS to perform some classification tasks on the
KDD98 Cup targeted mailing dataset. The dataset has nearly 100 000 rows and
about 400 attributes.
I create a mining structure based on a named query and a mining model using

i'm using MAR to discover association rules in a model, my model contains more than three nested tables, when i try more than two "input"s columns, then sql server doesn't proceed my model and it hangs with this message "Lecture de cas?en cours?:?0 cas lus." which means "Reading ...

in the time series model i want to predict for each hour for the comming year (arrival data) far a simulation software, how can i cache the data that the predict function would return an immediate answer and not take a long time?
(when i did a prediction as so:"SELECT flattened ...

if i have a model that has arrivals spreading on two years aggregated by an hour meaning 24 points a day and 24*7 a week. what would be the correct parameters for the time series historic_model_count, gap and hint????