I was surprized when SQL Server 2005 produced a desision tree with one branch referring to the same variable at more than one level. E.g it produced branches like these: All--A>100--B>10--A>120 All--A>100--B>10--A=<120 All--A>=100--B<=10 All--A<100 instead of otherwice expected: All--...more >>

Hi group, processing OLAP based mining structures "Errors and Warnings from Response" gives me the error message: Internal Error: Unexpected error (File 'dmcorrcountutils.cpp', line 2045, DMCC_WorkerJobCaseProcessor::CacheCase-Funktion) and the process fails. I am using SP1 with the last...more >>

How do I save the prediction query results to a relational table outside of the BI studio?...more >>

I was walking through the Text Mining example - which at one step required me to set Algorithm Parameters - MAXIMUM_OUTPUT_ATTRIBUTES=0. When I tried that the project would not build giving an error - Error (Data mining): The 'MAXIMUM_INPUT_ATTRIBUTES' data mining parameter is not valid for the 'XY...more >>

Hello! I have a question about mining structure. I read that data mining structure "holds" compressed data. That is useful for processing mining models while data don t need to be queried again from database... What is the upper limit of data which can be cached inside mining structure? A few mil...more >>

A related question: I have a MS Clustering model based on knowledge-area and I want to rename each cluster to the name of the knowledge area. What would be the DMX to rename the cluster? I have been trying to do something like - UPDATE [Knowledge Base].Content SET Node_Caption = [Knowledge...more >>

I've been working thru the Data Mining Tutorial (in order from the beginning) quite successfully until I reached the Lift Chart section. I get the following error message when I click on the Lift Chart tab: Failed to execute the query due to the following error: XML for Analysis parser: The XM...more >>

Hello! I am comparing MS Decicion Trees algorithm (classification) to Weka J48 algorithm. I used the same dataset for both algorithms. J48 uses information gain(enthropy) method for tree building. I also set parameter (SCORE_METHOD) of MS Decision Trees algorithm to Enthropy(1) method. Is this Et...more >>

Hi, As a beginner with DMX language, I just did the Bike Buyer DMX Tutorial. When it comes to Lesson 5: Executing Prediction Queries, the batch prediction query show the results, but how can I save the results in a physical table ? Thanks!...more >>

I am working with a sequence clustering model and trying to get my queries to match the information I see in the Model Viewer. In the model viewer I have State 1118 going to 438 with a 50% probability. State 341 also goes to 438 with a 43% probability. When I execute the following query: S...more >>

We have our primary application in SQL Server. Our end users are really struggling because they used to rely on Access to do a lot of their data mining and analysis. Our database has become so robust that Access is not really an option any more. Can anyone let me know if they've encountered...more >>

Hi. I have the following scenario.. TABLE1= sale.accom TABLE2= sale.meals TABLE3= sale.other I would like to get the Sum(price) of all 3 table and match against a client. I would just like to know the official way of doing this before I implement something that works in some scenario...more >>

I'm trying to evaluate SQL Server 2005 data mining in order to make a = potential purchase decision, but it keeps crashing when I try to view = the results. I have a table of user-generated events, and I'm trying to = determine the most common patterns of event sequences. Each event has = an e...more >>

Hi all. I got the next very simple and strange problem. When running the following 2 slqs, one runs very fast, and one very slow. I looked at my indexes time and time again, and cannot figure out why. The table has about 8 million rows. Very fast: Select top 1 keyA from tableA where fiel...more >>

I try to find a project about datamining, using OLEDB for DM (API) and SQL server 2005, simalar to Angoss SDK (using OLEDB for DM and SQL server 2000). Please help me. Thanks....more >>

I am experimenting with a program that is generating several thousand decision tree models and training them for me. My problem is that the performance starts dropping off severely once I have that many models on the server. For instance, when I try to open a DMX query connection it takes about ...more >>

hi When I built the same time series model using both DMX and BIDS based on cubes, I got very different results. Now I have a guess and I want to make sure my doubt is right. My data in cube is every day sales data, so for Sat & Sun, these data should be null. However in BIDS, the DM...more >>

Hi, I’ve created and trained a Time Series Model with Predict variable Amount and a Category variable which has the two categories ‘Junior’ and ‘Classic’. The values for Amount have a seasonal pattern for weeks, so I adjusted the Periodicity_Hint to 7. The results of the model are reasonable and in...more >>

Hi all. I am running into a problem with creating views that query between dates. I am attempting to search beween months regardless of the year. Whenever I create the criteria using: SELECT JobDescription, ProjectName, EstimatedFinishDate FROM dbo.BB_vw_ScheduledStarts WHE...more >>

Hi! I'm working on a plugin algorithm, that needs a method to somehow set parameters AFTER the training. I mean, I'd like to train the model with a set of parameters. Then, I want to set an other set of the parameters, and query the results. Then set this last parameter set to some other value...more >>

Hi, I am writing an application where i have a job which creates Mining Model Structure and its respective Minig Model. This job can be called by multiple threads at same time, thus creating multile MiningStructures and their mining models. Each of thread does structure.Update and model.Proce...more >>

I'm trying to wrap my head around the Time Series model, and so far am not having any luck. I'm reading through "Data Mining with SQL Server 2005" by MacLennan and Tang as well as the tutorial but I haven't had the breakthrough yet where I can say I really understand what is going on. I have a la...more >>

hi I want to add an mdx filter to the mining structure data source in my C# code. I use MiningStructure.FilterComponent but it did not work. I saw in XMLA, my MDX filter in FilterComponent only appears as child of <Annotation>field. However in other models I saw correct filter should appe...more >>

Hi everyone, I have downloaded the DM Sample Application, in order to query a Data Mining Model I have developed. Everything works just fine. My only problem is that I can't find a way to export the DMX query results out of the DM Sample Application (e.g to an excel spreadsheet). Is there a way t...more >>

We are just starting to get into the 2k5 mining tools. (and are very eager to keep going). We can create a project, we can run a project, however, as soon as we want to view the results in the mining model viewer (or any other viewer for that matter) we get an error. For the microsoft tree viewer, ...more >>

hi, When I was testing tips and tricks 1356 (Code and Reporting Services project from DM-BI Integration Webcast), I met a problem. For the following statement: Utilities.CreateMiningStructureColumn() I can not find the references anywhere. I do not know where is that Utilities ...more >>

hi, When I use the following statement to add cube slice filter into my data mining model MiningStructure ms = ... ms.Source = ... ms.FilterComponent = "Select ([date].&[20000101]:[date].&[20010101]) on columns from [mycube]"; The filter did not work and the model still too...more >>

Hi! I'm working on a clustering plugin. Everything seems to work fine, but when I call "SELECT Cluster()..." from the management studio, it says: ----------------------------------------- Executing the query ... Obtained object of type: Microsoft.AnalysisServices.AdomdClient.AdomdDataReader ...more >>

hi, When I am using AMO to create mining structure as: MiningStructure ms = db.MiningStructures.Add(...) ms.Source = new DataSourceViewBinding(dsv.ID); ms.CaseTableName = dsv.Schema.Tables[0].ToString(); I want to use cube data as my training data. Would you please...more >>

hi, When I am using C# for data mining, I found sth confusing. I can use ADOMD to create mining model like: AdomdCommand cmd = ... cmd.CommandText = "CREATE MINING MODEL ... cmd.ExecuteNonQuery(); However I can also use AMO to create mining model like: MiningStructure ms...more >>

Hi When looking at the template of the calculated member [Percentage of Total] it appears that I have to refer to the <<Target Hierarchy>>. That means that I have to create a calculated member for every hierarchy. Is there a way to create only one calculated member that refer to current hierarc...more >>

If you want to create a ds-tree using one continious input variable and a few other descrete, it would be good in some future version of analysis services to be able to have a column modelling flag like "REGRESSOR_ONLY". I.e. making it possible to stop the algorithm from using this input variable to...more >>

Why, when and how should ordered, cyclical, probability, variance etc be used as Data Type Content for a model column? Is this documented somewhere? Smith...more >>

DsTree parameter SPLIT_METHOD creates quite different results using 1,2 or 3. Are there any rule when to use which method? EBG...more >>

How appropriate is it to use SQL data analysis Naive Bayes classifier as a spam detection engine? I've been looking for a number of different solutions for a new product we're building and since its foundation is SQL 2005, I wonder if it does not make sense to make use of the Naive Bayyes algorit...more >>

Hi, I am trying to create a Mining Structure and Model using AMO library. I capture the XML Log to see what XMLA script gets send to the server when i do structure.Update() and model.Update(). The script seems correct but i never found structure being created on the server, if i seperatly take...more >>

Normally Decision trees and neural networks are the best algorithms to predict continuous variables, based on input of discrete attributes. But a Cluster can also predict a value. And if having say 3 million records, 10 discrete input attributes, and one value we want to predict, the cluster is not ...more >>

I have a time serie model built with multiple series using ”predict only”. Works pretty good and you can see how the model extends the series into the future. But I know that there are clear correlation between many of the series in my model. So when I change the model to “predict”, and let the m...more >>

is there any way to see a field of the standart deviation of the data when using the predicttimeseries prediction query?...more >>

Hi All. I am new to datamining and mssql2000. I bought a book called "Data Mining with SQL Server 2000 Technical Reference " I am not having to much luck with it as there is no database for the samples. I also cant seem to get the analysis services working on my machine, it cannot find the descis...more >>

hi, I use the related codes in AMOSampleDM.cs in Downloads to create my own Clustering mining model through Visual Studio. And then I use BIDS to create the another model with the same mining settings. I did not change any parameter settings so they both use the default settings. I use...more >>

Hey guys, I am looking to clean up (check for blanks, nulls, duplicates, etc...) some tables using the new features of SQL 2005/Data Mining. Any promptly suggestions are greatly appreciated. Regards, Jaqueline...more >>

Hey guys, I am looking to clean up (check for blanks, nulls, duplicates, etc...) some tables using the new features of SQL 2005/Data Mining. Any promptly suggestions are greatly appreciated. Regards, Jaqueline...more >>

I have some accounting data, with some transaction attributes and amounts. I'm using Decision Trees to try and predict the next month's amount for certain combinations of attributes. I've tried two different structures for the model: A: one with 9 discrete text input attributes. B: And anoth...more >>

Hi, What is the approach to get the leaf nodes of a decision tree? You allow us to drill through on the Decision Tree leaf nodes which means there should be a way to link the node to individual cases. I would like to use this as part of the prediction query to know why a particular case is cl...more >>

Hi, I am using SQL Server 2000. I have created one database named : ICS This database was working fine just few minutes before one of its table named "LogInOut" has stoped responding. Now even if i go for deletion it doesn't respond. Rest of the tables are responding normally. What could be...more >>

Does anyboby known the way that Occam's razor of Microsoft Decision Trees works? What is the rule that it uses? How makes the decisions for the pre-pruning? Any contribution will be helpfull...more >>

Hello! I found out a little strange behaviour of the PushCase(MiningCase inputCase) method for handling NULL values of the input cases. For example input case has a form: (Attr1 - 23.5, Attr2 - null, Attr3 - 11.3, Attr4 - null) I used iterator to iterate over the MiningCase object, and all t...more >>

I am looking into SQL 2005 analysis manager. Prior to applying the software, can anyone point me to software that can preprocess a recordset of opinions , and deduce such attributes as gender, age etc.? Thanks Jack ...more >>

Does MS SQL Server 2005 STANDARD Edition, contain the component "Analysis Manager"? -- Rob "Eddies in the space/time continuum" Ford Prefect...more >>