Topics: Coronavirus | AI | Data Science | Deep Learning | Machine Learning | Python | R | Statistics

KDnuggets Home » News :: 2013 :: Jul :: Competitions :: CrowdAnalytix: Predict the next phone call using RandomForests ( 13:n17 )

CrowdAnalytix: Predict the next phone call using RandomForests

The goal of this contest is creating right set of features to model and predict a users next phone call, using only RandomForests. Submission deadline July 25.

Predict the next phone callMODELING: Predict the next phone call

One of the business problems that offers immense potential for the telecom companies is to predict behavior of users and use that to market offers and services.

Predict the circle to which a user will make his/her next call. This objective is expected to be accomplished in two steps

  • Step1: from the given RAW data, you are expected to make features which are representative of the modeling objective i.e. to find which circle is the user going to call next? A minimum of 10 new features are expected as deliverable, such as frequency of calls, call length, time factor etc. and to be used as an input to model.
  • Step2: Solvers would use the features created in Step1 to build an optimal model for predicting the circle the user will call next on the given test data

Twist: The model is fixed for this contest based on client requirements - You can ONLY use RandomForest as your modeling technique.

Data: The data is anonymized user call log data for approximately 74K customers calling within various circles in the network. There are approx. 410K records in total.

Submission Deadline: 25 Jul 2013

Prize: USD $ 2000

For more information, and to participate visit

Sign Up

By subscribing you accept KDnuggets Privacy Policy