Training Course Overview

Our advanced data analytics training courses provide an in-depth insight into predictive modelling, churn analysis and credit risk models. Participants will acquire knowledge and skills required for planning, development, implementation and monitoring of analytical predictive models in the telecom industry. 

Key Training Topics Include:
	• Overview of Data Analytic and Predictive Modelling  • Data Requirements for Model Building • Advanced Modelling Techniques   • Risk Areas and Methodologies • Machine Learning Techniques  • Model Implementation and Monitoring

Analytics training is available in 3 distinct training courses:

– Telecom Data Analytics and Predictive Modelling Training: this is an advanced analytics course focused on practical data analytics and general predictive modelling.

– Telecom Churn Modelling Training Course: advanced course specifically designed and focused on predictive churn analysis and development of churn models.

– Telecom Credit Risk Modelling Training Course: course created for credit risk modelling with aim to increase the probability of payment result of post-paid customers.

Course Modules Breakdown (Example 2 Day Course)

Module	Short Description
Day 1
Market Definition	Key questions to ask in order to define the problem and modelling requirements
Application Process	Define key business processes where modelling will intersect with other systems
Account Management	Marketing and financial implications of predictive modelling and impact on customers
Outcome/ Response Definition	Define required outcomes, criteria, time limitations, volumes of data.
Data Requirements for Model Building	Availability, complexity, requirements. History and applications data. Internal & external data. Customer demographics, product and billing data.
Introduction to Modelling Techniques	Overview of univariate analysis, correlation analysis, variable reduction techniques and statistical regression
Day 2
Advanced  Modelling Techniques	Univariate Analysis – Catalogue of indicators, data quality (missing, invalid values, improbable values), predictive power, stability analysis Variable Reduction Techniques – Correlation analysis, co- linearity Regression – Clustering for market segmentation – Inference for declined profiles – Decision trees – Survival analysis techniques
Machine Learning Techniques	– Where it is used and description of key techniques – Explainability and interpretation difficulties, over-fitting concerns  – Neural networks  – Random forest  – Genetic algorithms  – Predictive text mining  – Deep Learning  – Ensemble Models
Strategy Design and Using Your Model	How to use your model, how to set cut offs, risk based pricing, rule setting, cross sell/up sell/ down sell
Model Monitoring	What to monitor – model characteristics, performance, stability, strategy adherence.