So Octave Analytics is inviting you to attend a 1-day Business Analytics and Machine Learning Master class that will hold between January 18th - 20th 2018.
Admission is free. Due to space limitation, participation will require registration.
Venue: Board Room, 6th Floor, ccHUB, 294 Herbert Macaulay, Yaba, Lagos.
The focus will be Churn Prediction Model - A Banking Case Study.
Problems:A Bank (not Guarantee Trust Bank) is looking for help from data scientists to help provide insights using its past data. The bank wants to predict the propensity of its customers to churn. This would help the bank to determine the right engagement or intervention plan.In a nutshell, the bank wants to identify the customers who are likely to churn between the next 91 - 180 days. It has provided its customers’ information such as age, gender, demographics along with their assets, liabilities and transactions history. The task is determine the propensity to churn for each customer. The bank provided 500,000 dataset of their customers, together with their various relationships along all the products lines. We have 300,000 customers data in training set and 200,000 in the test set. We are expected to estimate the probability for each customer in test set to churn. Due to the unbalanced nature of the data, Accuracy will not be the evaluation criteria. The evaluation metric is the percentage of responders (churner) captured in the first and second deciles from all the responders.
culled from https://datahack.analyticsvidhya.com/contest/data-science-hackathon-churn-prediction/