Use your predictive models better

Most statistical studies are done from representative samples of the studied populations. This helps reduce the processing time during development of different models, without affecting the validity of results.

Nevertheless, in most study cases carried out in enterprises, the development of the model is not the final objective. The objective is to apply this model to the entire target population. For example, to calculate an appetence score for each client appearing in a customer database or to segment all the individuals from a prospects database into different characteristic groups. Hence, the model is really useful only if it applies to the entire population.

Coheris SPAD lets you automatically apply a model established on a sample on the entire database. The model in question is only a succession of processings (conversions, calculations etc.). Once defined during the data management and modeling phases, with the help of the workshop and graphics tools, these processings can be reapplied to the data.

This generalized application of a data mining model can be planned and executed like a batch process on an adequate server.

Strenghts

• Automatic application on the complete target database, of the model created as sample
• Possibility to start these processings in batch, on a dedicated server.

Formations Coheris SPAD data mining