Welcome :Guest
Congratulations!!!

Top 5 Contributors of the Month
ffttt

### Decision Trees - significance of Node_Description features

Let's say that I have one node in my decision tree with a Node_Description that looks like that:

InputParameterA >= 10 and InputParameterB not = 'True' and InputParameterC < 10 and InputParameterD >= 120 and < 150

Is there a way to know what is the importance of each of the four input parameters?

For example, a potential answer might be "InputParameterA" contributed 32%, "InputParameterB" contributed 2%, "InputParameterC" contributed 10%, and "InputParameterD" 56%. Such an answer will clarify that "InputParameterA" and "InputParameterD" are more important.

Thanks!

### Learning Bayesian Networks for building Decision Trees

In MSDN article Microsoft Decision Trees Algorithm Technical Reference is said "The Microsoft Decision Trees algorithm learns Bayesian networks..". It seems to me that BN and DT are very different tasks and approaches to data modelling. The question is: how learned Bayesian network (which models joint distribution considering some conditional independencies) is used further to build decision tree?

### Decision tree dependency measure

The dependency network viewer executes the stored procedure System.Microsoft.AnalysisServices.System.DataMining.DecisionTreesDepNet.DTGetNodeGraph which yields some integer measure that represents the strength of influence of input variables on the output variable. How this measures are calculated (information gain, chi-square, correlation etc)? It seems as if conditional influences are not taken into account: if A and B are two major factors which impact variable C, but A and B are strongly correlated so that C given A is not dependent on B, dependency algorithm will still depict B as the second major factor. Am I right? So there is no analogy of tests like conditional chi-square, conditional mutual information or partial correlation?   It seems to me that MS Data Mining lacks some kind of Bayesian Networks algorithm wich would illustrate conditional dependencies. That would give a useful insight on how various factors are related to each other and through what kind of chains a change in some input variable transforms to the output.   Thank you.

### SQL Server 2008 performance: Decision to spend budget to upgrade to enterprise edition

Hi, Currently, we use SQL Server 2008 standard edition. We are planning to improve the performance of our application. The application is built by 3<sup>rd</sup> party and we can’t change/modify database objects and they seem to be designed fine. The options are spending our budget to upgrade to sql server hardware, or upgrade to SQL Server Enterprise edition. I am a bit confused over the benefits of SQL Server Enterprise edition. The following chart suggests that SQL EE includes “Parallel Index Operations” and “Enhanced Read-ahead and Scan”: http://www.microsoft.com/sqlserver/2008/en/us/editions-compare.aspx  How significant are “Parallel Index Operations” and “Enhanced Read-ahead and Scan”? Considering the fact that new hardware is cheaper than sql enterprise license, Should I spend budget to upgrade enterprise edition instead of buying as server with faster CPU and RAM? Thank you, Max

### CASES & CONTENT of a Decision Tree Mining Model in One View

I would like to have a SQL view that presents both the CASES & CONTENT of the decision tree mining model in one view. Meaning - the view shall consist of the following fields: Node_Name, The Column to Predict, Unique Number (or Key) of a case, and input features of the case. Here is an example:

Node_Name        | Key |   Column to Predict   | Input1 | Input2 | Input3 |...

0000000000101      |57|

0000000000101     |372|

0000000000100     |118|

0000000000100      |52|

0000000000100     |623|

000000000010000  |41|

000000000010000 |205|

I'm familiar with queries such as those:

SELECT *
FROM OPENQUERY(AnalysisServer,
'SELECT * FROM [MyMiningModel].Cases
')
As DMX

SELECT *
FROM OPENQUERY(AnalysisServer,
'SELECT
NODE_NAME
FROM
[MyMiningModel].Content')
AS Dmx

I wish that there was a way to present the information in one view.

Thanks!

### How to assign 'Maximum Input Attribute' parameter for Decision Tree in SQL Sever 2005 in vba

Hello,

I'm trying to assign the 'Maximum Input Attribute' parameter of the Decision Tree algorithm during forming a Mining Model in vba (see the code below). But I get an error message, that it is invalid for my Decision Tree. Is this parameter available only for bayesian networks or it appears only in version 2008 or am I doing something wrong?  Thank you very much in advance!

SetAlgorithmParameters()
...

Valentina

### How do you know which model is best eg: C4.5, k - Nearest Neighbour, and Decision Stamps

After finding the accuracy, error rate, sensitivity, specificity, precision for each model. How would you be able to tell which model is best model among the models

### EXTRACT IMAGE FROM JPEG FILE - LIKE AN EAR, MOUTH, EYE, OR ENITRE FACE MAYBE I JUST WANT THE TREES

I want to be able to pull all separate images out of a picture and do this via VS 20120 maybe VB, C++, C#, .NET. CONSIDER the following:

An image file has a picture that contains 3 persons, 1 dogs, 1 cats, a bear, and several trees. I want to create 3 separate image files with 1 person, a tree, and an animal displayed in each image.

I can do this manually utilizing photo shop or other image processing software But, I want to do it through a program.  Why, well I just want to learn how to identify a specific frame in a photo and do whatever I want with that image file. I am unsure if the image file is made up of frames that isolate each image that makes up a photo?

It will be challenging and a LEARNING fun task to start from my point of knowledge concerning an image file (which is less then none)  and end up isolating individuals images that make up the story of the photo.

Any thoughts, source concepts, snippets, etc.. would be greatly

 Categories:
 ASP.Net Windows Application .NET Framework C# VB.Net ADO.Net Sql Server SharePoint Silverlight Others All