Subscribe to DSC Newsletter

Which software do you like best for business intelligence, fraud detection, web analytics, data mining, or risk management?

What do you think of SAS Enterprise miner? Which procedures do you use most? Do you use it mostly to process large datasets (more than 10MM rows, more than 100 variables)? If you use SAS/Stat, which procedures do you find most useful (please give context, e.g. pharmaceutical, small samples etc.)

What about open source software (R) and Splus, Salford Systems, JMP etc? My experience is that they process your entire data set in RAM, and are thus limited when dealing with large data sets (I could not get JMP decision trees to process more than 300,000 rows of summarized data). Did you manage to solve some problems thanks to Syncsort?

When looking for a vendor, comparative studies or benchmarking, what is your favorite reference?

Tags: R, SAS, datasets, large, open, software, source, statistical

Views: 101

Reply to This

Replies to This Discussion

In SPLUS you now have the choice to process your large data sets as BIG DATA, so not in RAM but of course the processing time is slower. But I guess given the flexibility you get with S language I find it very comparable to Eminer although not all predefined procedures in Eminer are readily available in SPLUS, so if you are a fan of predefined procedures rather than custom codes then I guess Eminer is the winner. But I rather go with custom coding as I find predefined procedures very rigid and inflexible.
I wasn't replying to your comments rather the main question posed by Vincent. Please see the main question of this discussion.
I would welcome comments- positive or negative - from anyone who has experience of commerciallya vailable packages - and for that matter of open source ones. We have used Clementine in the past - experience was wholly satisfactory, AnswerTree recently - it did the job well, and we are thinking of using RuleQuest. I would therefore welcome any comment from someone who has used RuleQuest in anger.

I'd also like to hear from anyone who ahs linked the machine learning and rule-based system paradigms. That is people who have developed working rule-based systems using procedures or rules generated by an ML program.

We have done it for early predicting of student attrition.

With thanks in advance

Laurie Moseley
University of Glamorgan

RSS

On Data Science Central

© 2019   AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service