Difference between models in Operators vs Auto Model

jsdrew

I am new to Rapid Miner and I'm trying to understanding the difference between the models in the Operators folders vs those used in Automodel.  I've been experimenting with the Deep Learning Predictive Model.  The one in the Operators folder seems to create a model from a dataset much faster than running the same dataset to create a model through Automodel.  

Thanks for your help on this.

MartinLiebig

Hi @jsdrew,

the techniques used in AutoModel and in Go are very similar, but not 100% the same. You should get very similar results on Go.

Let me help to differentiate Studio and Go:

What Studio is

* A full fledged visual workflow designer (operators). Including dozens of extensions

* The ability to connect to a RM AI Hub for deployment, collaboration and scaled out execution

* RapidMiner Turbo Prep for a 'guided' way of doing data analytics

* RapidMiner AutoModel for automatic modeling.

* Software on your laptop

What Go is:

* A SaaS offering for the cloud

* An AutoModel capability similar to RM AutoModel

* An option to deploy using WS on our cloud

* A more guided, explained experience (compared to RM AutoModel).

* Way cheaper

The RapidMiner SaaS offering will be extended in the future, so you can expect a lot of new things on the SaaS side. Go is basically the first (corner)stone.

I hope this helps?

Best,

Martin

MartinLiebig

Hi @jsdrew,

keep in mind that AutoModel does more than just fitting a model. It also does a ton of preprocessing for you. In principal you can do all the things auto model does manually. If you want to check what AM is doing, you can just open the process it creates for you.

Best,

Martin

jsdrew

Thank you for your comment @Mschmitz. I am also trying to determine the difference between Rapid Miner Go and Rapid Miner Studio in terms of modeling. If I run the same dataset through a deep learning model in Rapid Miner Go and auto model in Rapid Miner Studio, will I get similar results? While I've worked with databases and data for many years, I'm new to data mining so I don't know a lot about models. Besides the data prep capabilities, are there any other advantages of Rapid Miner Studio over Rapid Miner Go? And, in terms of the data lines limit for the Pro License for Rapid Miner Studio, does that apply even when you've downloaded the software to your own PC? 

I'm not sure if this is the right place for these types of questions, but the sales people I've talked with at Rapidminer so far haven't been able to help out with technical specifics like these.