Ask A Data Miner - 75,000+ Members
I am starting an open source data mining project in C# .NET. I would like to post the C# code on this site, then, have other people contribute to the development of the project.
If you are interested in participating, please reply to this thread indicating what data mining algorithm you would want to participate in.
Replies (50)
Hi ThereThanks for your reply. I have seen your C# project on this site and have already downloaded it and played with it. I have managed to import the project into Delphi 2005 and convert it to a Borland C# Project.My Problem is that I am unsure how to use the classes in your library and was wondering if you had a real world example application that implements your library and uses it to cluster a sample array of data or datatable.I have been trying unsuccessfully for about a week to write the algorithm in Delphi. I can get the initial clusters ok, but I need to optimize the code and figure out how to calculate SSB, SSW and Calinski index correctly. This is my first attempt at clustering and I have written my code based entirely on clustering documentation that I have found online.The Eculidian distance function is what I use in my code.I would like to use your code and my code to cluster my data and compare results whilst debugging. That way I can see where I am going wrong.My Data is an array with the following structure:Store Number, Sales Units, Sales Value, Store BaysI have up to 1000 rows of data in my array. I basically need to cluster stores for similarity purposes before I can use the forecasting engine I developed a few months ago. Problem is that some stores lack sales history (newly opened stores) and I need 24 months of history to forecast accurately, hence the need to cluster by similarity and use similar stores for forecastable values.Thanks again for your reply and I look forward to hearing from you again.Kind RegardsDave Pike(South Africa)
Kingsley, you can count me in! I know little about data mining,but I would like to contribute code for distributing search indexesacross multiple machines, where many machines could fail.
Is this an example of datamining? I hope so, because if it is, then I found my topic [:)]
[quote user=MadDog]Hi ThereThanks for your reply. I have seen your C# project on this site and have already downloaded it and played with it. I have managed to import the project into Delphi 2005 and convert it to a Borland C# Project.My Problem is that I am unsure how to use the classes in your library and was wondering if you had a real world example application that implements your library and uses it to cluster a sample array of data or datatable.I have been trying unsuccessfully for about a week to write the algorithm in Delphi. I can get the initial clusters ok, but I need to optimize the code and figure out how to calculate SSB, SSW and Calinski index correctly. This is my first attempt at clustering and I have written my code based entirely on clustering documentation that I have found online.The Eculidian distance function is what I use in my code.I would like to use your code and my code to cluster my data and compare results whilst debugging. That way I can see where I am going wrong.My Data is an array with the following structure:Store Number, Sales Units, Sales Value, Store BaysI have up to 1000 rows of data in my array. I basically need to cluster stores for similarity purposes before I can use the forecasting engine I developed a few months ago. Problem is that some stores lack sales history (newly opened stores) and I need 24 months of history to forecast accurately, hence the need to cluster by similarity and use similar stores for forecastable values.Thanks again for your reply and I look forward to hearing from you again.Kind RegardsDave Pike(South Africa)[/quote]
I'm a graduate student major in Data Mining in Taiwan.
This is our team web site crated by my teacher, but the language is chinese.
http://www.kdkeys.net/open-source-c-net-data-mining-project/#link-6861
Sorry,my english is not good enough .But I'm very enjoy your project and hope join with you.
My special skill is VB.NET.I know your topic is using C#.It doesn't matter, you know
.net framework which language is not important.
Could you count me in ?
Maybe G.A.,G.P.,SOM,Sequential pattern,Clustering, as well.
Hi
I am currently working on a final year project (B.Tech) .Topic: Modeling an Intrusion Detection using Data Mining. I have used Snort as Open source Intrusion detection system. Can you help me with some code for Data Mining Algorithms like K-Means, Decision Tree, Apriori, Learning Algorithms... Any sort of help is appreciated.
Please send me the available source code and ideas to knk1988@gmail.com
And thanks to all
Regards,
Naveen
Dear Satish Thumar :
Thanks for wanting to contribute your expertise in data mining.
Firstly can you post a very clear description or paper on neural networks.
Please take a look at some of the data mining algorithm descriptions at http://www.kdkeys.net/open-source-c-net-data-mining-project/#link-6862 e.g. at http://www.kdkeys.net/open-source-c-net-data-mining-project/#link-6862/archive/2004/03/21/169.aspx their is a description of a data mining algorithm with Naive Bayes and at http://www.kdkeys.net/open-source-c-net-data-mining-project/#link-6862/archive/2004/03/21/168.aspx a description of a simple One Rule Algorithm.
The goal of the documentation is to make it clear to everyone what the algorithm does and the programming logic that will be involved. All the mathematical calculations will have to be explained, etc.
Once this is complete, then the fun Coding part starts[:D]
The coding can be done completely by you with others testing, documenting, etc. or you can ask for as many participants as necessary. However nothing can be done without the documentation as no one else will understand what you or we are doing. The documentation is the blueprint or plan for the house and after the blue print is clear to everyone then we can build the house.
Will you be able to help us with this first part of the task : to create and post a detailed tutorial on Neural Networks?
Thanks,
[quote user="satish thumar"]HiI am doing a Ph.D in the field of data warehousing and data mining. My Aim is to develope a mining algorithm using a neural networks. So just send me the details of your need.Satish Thumar[/quote]Dear Satish Thumar :Here is what I hope is a Naive Bayes Data Mining Tutorial at http://www.kdkeys.net/open-source-c-net-data-mining-project/#link-6865. From this point one can start coding on Naive Bayes. We need a Tutorial like that for Neural Networks.Thanks,
Dear Germic :Which data mining algorithm or topic or any other area would you like to contribute on?Thanks for helping [:D]
K-MEANS ALGORITHM SOURCE CODEIMPLEMENTATIONINC# BETA 1 RELEASE AT http://www.kdkeys.net/ShowPost.aspx?PostID=3538
Hi ThereI am currently developing a k-means clustering system for clustering store related data (Sales units, Sales Values and Bay information). I am primarily a Delphi developer although I have prior experience with VB and some knowledge of C#.Would it be possible for you to give me an example (C# source is fine) of how to implement your k-means classes and use them to process a sample array of data. I would like to compare these results with my current implementation and optimize my code. (Having difficulty with SSB, SSW and Calinski index)I will be happy to share my code with anyone who is interested in a pascal implementation of k-means.Any help here would be tremendously appreciated.RegardsDave Pike(South Africa)
You can get the VB.NET and C# versions here
http://www.kdkeys.net/forums/60/ShowForum.aspx
[quote user="data_miners"]
Can i have a new version of Neural Network Backpropagation ??? where can i get it???Cause i want to compare the source with mine [/quote]
Hi, I m ShivaBrindha..I m donig my project in "Intrusion detection using datamining techniques"..I need source code for association rule mining using Genetic network programming and fuzzy logic... for classification i need the source code for c4.5 and SVM. Please help me..i need the source code in .net...Please sir.. send the code to "shiva.akash@gmail.com"...
Follow On Twitter
Request More Information
Hi :
Have you seen the open source K-Means C# Implementation on this site?
It is functional, not completely tested but it looks like it will meet your needs.
Could you describe the K-Means algorithm and techniques you implemented?
Also on this site, their is a link to a free source code Visual basic 6 kmeans algorithm implemenation.
Also, have you thought of getting your hand on a SQL Server 2005 Beta Version and using the Clustering Algorithm?
Please let me have your comments and thoughts along these or any other lines.
Thanks