Data mining / Auto Classification /Rules extraction

I've done music classification in both Java and Python before. Please let me know if this is a class project or a commercial project and what language, library and platform can/should be used.

Hello Sir, We are a well built set up with excellent skilled operator with lot of experience in this segment/skill,have complete more than 200 similar job, i have gone through your project description, its really a simple task, we can make a well formatted,error free list. Sir, we are waiting for your reply for communicate also like to make a sample if you allow by a single reply. Thanks DG

Dear I have just seen your Job posting and upon reading thoroughly, I felt strongly that I would be the most qualified contractor for your project. I have vast experience in IT / Data Entry / Web Research /Admin / SEO / Content upload for eCommerce and Marketing. Understanding how valuable an education is, I am serving Pacific Pharmaceuticals Limited as a Asstt: Manager (MIS), toward earning my Masters of Business Administration (MBA) degree and B. Sc Computer Science. I have completed many web research/data entry/ General Admin and marketing project. I am 100% able to do your project. Sincerely, Tarikul Islam

Hi, We are a group of analysts and working in the field of data processing, data mining and analytics domain. Below is our plan of project. We are assuming that the class variable you are mentioning is the target variable. It may be binary (1,0) or categorical or a continuous variable. The other 10 variables you have are explanatory variables. What we have understood is that you are trying to make few rule based segments based on these 10 variables. Distribution of class variable would have to be significantly different in each of these segments. To achieve this we can use an objective segmentation( tree) method like CHAID or CART. However to make the rules robust and stable there are few more steps we would have to take. As for example creating a development and validation sample from 10,000 records (may be a 70%, 30% split). At the end of the exercise suppose we get 5 different segments using 4 explanatory variables, we would put any new record in any of these segments based on the rules. At this stage if we get to see the data(few records in excel with variable names), it would be easier for us to furnish further details. There are some other mothods available. But that would completely depend on the end objective. Looking forward for your reply. Regards, Pal

HI Brother, I am Data Scientist working in Multinational Company. My work is to see the hidden pattern in the large and complex data sets and predictive analytics, Data mining,Machine Learning and also uses the statistical techniques on those datasets to see what data behaves in past. By seeing the past what will be the behavior the that data set in future. I am also working on "kaggle" competitions currently i am predicting the default loan prediction for imperial college London.I am also writing the column on All Analytics and my rank on All analytics in Statistician. I am also working on the Time Series analysisusing ARIMA, SARIMA models etc. I am also using R, SPSS Statistics, Spss Modler, Stata, SAS, MATLAB, Statistica and Python for Data Science for Machine Learning, Data Mining and Data Science.I have also several publications relating Data Mining one which is related to you project is "Data Mining in Insurance Claims(DMICS) two-way mining for extreme values" and you can sight this paper at IEEEXPLORE. Thanks Regards, Muhammad Shoaib. Data Scientist 360 Technologies,360 Logics Department, Islamabad, Pakistan.

I think the problem would be solve by a statistical method call principal components or cluster analysis using statistical software like SPSS, stata or eviews. As expirience I can mention 2 years in survey data analysis and data mining support to social scientists.