Methods for learning from Telco Data

In the previous section, we described our use case of learning customer insights from big Telco Data with a new dynamic approach and also prepared our Spark computing platform with SPSS on Spark and MLlib as the focus. By following the same process adopted in the previous chapters, as the next step for our machine learning, we now need to complete the task of mapping our use case to machine learning methods. We need to select our analytical methods or predictive models (equations) for this project of scoring customers with big data on Spark. Even during our machine learning, we may need to jump back to this stage. We shall still fully complete this stage and prepare all the needed knowledge and codes for jumping back here easily during our machine learning process.

As for the modeling part of our learning, that is, to model customer behavior with Big Data, which is either to depart or to call services, or to purchase for our case here, there are many suitable...