2. Are end users required to be knowledgeable of statistics?
A: No prior training or knowledge of statistics is required. Users simply enter the information and Decision Science returns the answers. All advanced analytics occur behind the scenes, with minimal user intervention required.

3. In what programming language is the SDK written?
A: The SDK is currently provided as a C/C++ API. The functions are easy to understand and there are many examples of sample code in the documentation to guide the integration process. Other language interfaces will be provided in future releases.

4. Does the SDK work with a Microsoft® Visual Basic® application?
A: Yes. The SDK is Visual Basic safe, and the same API can be used in your VB applications.

5. Does the SDK work on Linux?
A: The current SDK works primarily on the Microsoft Windows® platforms. The code can be ported to a Unix environment based on customer interest.

6. How fast does the SDK build and deploy a model?
A: It depends on the number of records in a particular database, the database server utilized, and the network and hardware configuration.

7. What are the minimum resources needed to run an SDK?
A: It depends on the dataset being scored. 256 MB of RAM and at least a Pentium III processor are the minimum resources recommended.

8. What type of support is offered?
A: The SDK includes a developers guide, sample code, and technical support is provided via email. Contact a Stone Analytics Representative for information regarding more advanced technical support packages currently available.

9. How much time is required to integrate an SDK into my application?
A: This depends greatly on the application into which the SDK is being integrated. The driving of the SDK is straightforward; the level of interaction by the end user, however, will govern the level of effort required to integrate. For a basic, procedural application, the SDK can most likely be integrated within a few days.

10. What special pre-processing needs to be done to my data source?
A: The database should be relatively clean. The SDK manages missing and out-of-range data. But it builds a better model with superior deployment and faster performance if the database:

• Has no missing elements;
• Contains data consistent with field type and format;
• Has minimal outlier values.

11. What databases does the SDK support?
A: The SDK currently supports all ODBC-compliant database interfaces.

12. How do I obtain an evaluation version of either Scoring Science or a Valuation Science?
A: An evaluation version of the SDK can be obtained by contacting sales@stoneanalytics.com.

13. How is the evaluation version of the SDK limited?
A: The evaluation version is fully functional except that all model-building results are zeroed out, so that any output is essentially meaningless. The evaluation SDK will, however, provide you with the opportunity to integrate the product and determine its performance and characteristics within your application.

14. How can I purchase an SDK?
A: Please contact sales@stoneanalytics.com for pricing and purchasing information.