Expert-to-Expert Data and Text Mining Workshops

Duration: 2.0 hours

This outline is provisional and subject to change.

Do you want to leverage analytics, and are you asking yourself, "How do I get started or approach the business problem?" Do you have a specific business objective - and the data - but you want to know how to efficiently connect the dots from data to solving business problems? During this two-hour workshop, a SAS Analytical Consultant will work individually with you so you have a chance to think through things, get some hands-on experience, and learn step-by-step procedures, best practices, and tips and tricks for the analytics process - from exploratory analysis to predictive modeling and implementation. This expert-to-expert workshop will be conducted by top-notch Analytical Consultants who every day work hand-in-hand with customers to solve business problems using data mining and predictive modeling. Depending on your specific needs, the workshop can be geared toward approaches to accommodating missing data, using sampling guidelines, selecting variables, transforming variables, performing text mining, working with decision trees versus neural networks, and selecting the best model, as examples.

As part of this one-on-one workshop, you can bring your data that you want to analyze. However, if you are not able to bring your own data, we will have a variety of sample data that can be used to mimic your business situation. If you choose to bring your own data, please keep in mind that the duration of the session will be limited to two hours. Depending on the state of the data, it could take the entire two hours to simply process your data in preparation for data mining or modeling (for example, merging multiple files or summarizing records). Therefore, the more that you can do prior to the workshop to prepare the data for modeling, the higher the chances that you will get more out of the workshop. Please pay careful consideration to these suggestions:

• Any required merging or summarizing of data should be done beforehand.
• Modeling is best done on totally denormalized data.
• Having your data st ored as a SAS data set is preferable.
• Large volumes of data (over 1 gigabyte) could prove restrictive in terms of processing time.

Who should attend: This expert-to-expert workshop is designed for the executive, data analyst, business analyst, and/or statistician who wants to discuss approaches to solving specific business and technical problems related to analytics. Whether you are an executive who wants to understand how you can leverage analytics to outperform your competition, or you are in the middle of a data mining project and need some specific tips and tricks, or maybe you have already built a predictive model and need to walk through the process for building an improved version, this workshop will be customized to address your specific analytical needs. It is designed to be a "one-on-one" session for up to six individuals within your company who all share a common analytical need.

Prerequisites

There are no formal prerequisites for this workshop, and you do not need to be a statistician to attend this session. However, we do require that you complete the following items prior to the workshop:

• Prepare a brief write-up of your business objectives for analytics.
• Have notes that generally define the data structures that you might use for your analyses.
• Have notes that speak to the type of expertise that you have available on the business, data, and analytic levels.
• Complete a one-hour pre-workshop call with an Analytics Consultant.

Upon registering for this workshop and prior to the conference, a SAS Analytics Consultant will contact you to review your needs and to answer any questions about data preparation work you may have if you choose to bring your own data.

Course Contents

The one-on-one workshop will be geared toward your specific needs. Specific areas of focus can include:

• Discussion of approaches to solving business problems using analytics
• How to leverage your data to solving business problems
• How to deploy an analytic result in your operational environment
• How to work with the various players in your organization to most effectively solve an analytic problem
• How to divide a large problem into smaller more manageable units that can each be completed in an acceptable time period
• Topics specific to an analytic project, which may include the following:
  • Preparing data for the modeling process
  • Sampling, partitioning, performing graphical analysis, performing association analysis, clustering, filtering, and dealing with missing values
  • Variable selections, regression modeling, decision trees, neural networks, and model selection
  • Validating models, interpreting model results, scoring, and translating model output into actionable outcomes
  • Text mining

This page was created using SAS software.