An analytical mindset

For those readers who are analytics professionals, let's draw an analogy. Advanced analytics models are trained on data. The data represents the world or the subject area at the time that the data was collected. Once the model is trained and that model is accurately "predicting" the characteristics of the subject area as represented by the training data, the model is "locked." By locking the model, we end the training phase and we move the model into production.

The model ingests and examines data in the operational world and predicts the information that we are interested in. But we all know that the world changes and so does the data that is the byproduct of those activities. The models must be updated or retrained using current data to ensure that the models are generating predictions based on data that is as close to the current state of the world as possible. We "unlock" the model and train it again using new data....