Utilize data mining techniques (e.g. artificial neural networks, genetic algorithms, gbm, etc.) to develop a generalized model to help discern which students have the highest probability of success. The model would be able to be customized for a given institution using that institution's own historical data.
This is coming from the discussion by UMBC on Day 1 of #EDUSprint. Hrabowski talks about needing options to show people the benefits of decision making with analytics. In order to do this we need a culture of trust around the use of the analytics. We also need this data to work on a model that does not require mediation in order to make informated decisions.
Analytics, like computers, make good servants but bad masters. How can we use them thoughtfully and productively? Right now, they're cresting the "peak of unrealistic expectations" on the Hype Cycle! (See http://en.wikipedia.org/wiki/Hype_cycle)
What is a good first project for analytics? What are the top five reports to start with?
Analytics will result in reductions in force
The higher education community doesn’t know how to use data to make decisions.
A major benefit of analytics is to optimize use of resources.
EDUCAUSE current working definition of analytics is: Analytics is the use of data, statistical analysis, and explanatory and predictive models to gain insights and act on complex issues. Does this resonate with your understanding of analytics?
An important concern in the growing use of analytics is misuse of data
Containing or lowering costs is an important benefit of analytics.
Social Web