What inputs are required for the AutoClassifier product?

We require a modeling data set in a tabular format, and identification of the target field.

  1. The modeling data set can be defined as a file (table) with one row for each modeling unit.
  2. A modeling unit can be a ‘customer’ if you want to build a statistical model to make predictions at a customer level. Or, it can be a ‘prospect’ if you want to build a model to make predictions at a prospect level.
  3. Each column in this file represents a set of attributes for each modeling unit. Examples of modeling attributes are: revenue, age, number of transactions.
  4. One of the columns must contain the modeling target, and it must contain binary values such as 0/1, or Y/N.

We prefer the modeling dataset in a CSV format, but we can work with other formats as well; just let us know. When you upload the modeling data set on our secured web-site, we will ask you to denote the target field.

And that's all we need!

How long does it take before I can view my results?
Once you upload a modeling data set and specify the target field, you will receive an email within 24 hours or less when your results are ready to be viewed and exported from our secured web-site.
Is my data secure?

Yes. We use a secure and sophisticated cloud infrastructure to protect your data and information while they are transferred, temporarily stored, and processed.

How would I interpret and use the model output?
The model results are intuitive and easy to interpret. However, we will provide a brief guide that you can use to interpret model results and derive insights. If you need help with interpreting results, we’d be happy to answer them via email or phone.
How much does it cost to build a model using AutoClassifier?
Please check out the Pricing page.
Are there any limitations to the number of records, attributes, or the file size?
We are equipped to handle large files, number of records, and attributes. If the file size is prohibitively large, please let us know in advance so that we can ensure a seamless operation.
What type of scoring program is generated from AutoClassifier
Currently, the scoring program is available in Python and/or SQL. We are working to make the scoring program available in additional languages -- such as SAS, R.
Can I score another dataset in future by using the a previously built model?
Once our engagement is complete, we safely and securely wipe out all data from our secure servers. However, if would like us to keep the ‘scoring program’ for possible future use, please let us know in advance and we will securely store the scoring program for future use. Note that we will not store any of your actual data unless you specifically ask us to.
What type of support, and other services, do you provide?
Please check out the Services & Support section.