Core Content: Predictive & Prescriptive Analytics / AI for Business

Methodology

Methodological courses are an important part of this advanced master program. These include: Machine Learning, Deep Learning, AI, and predictive and prescriptive analytics. In an era where data is often called the "new oil," our program ensures you aren't just a refiner, but a master architect of the engines that run on it. Our curriculum is built on a rigorous methodological foundation that bridges the gap between theoretical mathematical constructs and real-world industrial impact. We don't just teach you how to use tools; we teach you how to build, critique, and evolve them.

AI

The program begins by demystifying the hierarchy of modern intelligence. While Artificial Intelligence (AI) serves as our broad North Star, we dive deep into the specific algorithmic frameworks that make it functional.

Machine Learning (ML): You will master the statistical learning theories that allow systems to identify patterns and make decisions with minimal human intervention. We cover everything from classical regression and clustering to ensemble methods.

Deep Learning (DL): As we move into high-dimensional data like images and natural language, we explore neural networks. Our labs focus on the architecture of CNNs, Transformers (LLMs: large-language models), and Generative models, ensuring you understand the "black box" rather than just looking at it.

Turning Insight into Action: Predictive & Prescriptive Analytics

Data is historical, but its value is inherently future-facing. Our methodology emphasizes the transition from understanding what happened to determining what should be done. Predictive Analytics: We focus on the "What's next?" Using time-series analysis and probabilistic modeling, students learn to forecast trends and identify risks before they manifest. Prescriptive Analytics: This is where true leadership happens. Beyond simple prediction, we explore optimization and simulation techniques to suggest the best course of action. It’s about moving from "It might rain" to "You should carry an umbrella and take the 9:00 AM train."

Domain knowledge

In an increasingly competitive world, just competing on superior product performance has become very tough. Therefore, companies have turned toward leveraging data across their entire business. Actively managing customer relationships is an important part of that. It includes the following objectives:

1. Acquisition (identifying & attracting new customers)

2. Cross/up-selling (profitable usage stimulation)

3. Retention (identifying customers who intend to attrite/churn, and trying to keep profitable customers)

4. Recapturing lost customers

Each of these objectives can be supported by analytical tools powered by traditional statistical techniques or data mining algorithms. Hence, the field of analytical Customer Relationship Management (aCRM) has seen stellar growth.

This new approach to conducting business has been acknowledged by book authors such as Thomas Davenport & Jeanne Harris in 'Competing on Analytics' (2007) and Ian Ayres in 'Super Crunchers' (2007). The DS4B program is not a master in marketing management, but it focuses on research and highlights mostly quantitative issues. The target group consists of both people with working experience and young graduates who feel the need for an in-depth education in marketing analysis. By bringing together a group of motivated students and teachers, and by dynamic and multimedia teaching methods, the Department of Marketing, Innovation and Organization is striving to transfer in-depth knowledge of the marketing field.

Analytical Customer Relationship Management

In analytical CRM, we try to capture customer dynamics, i.e. customer inflows as well as outflows. CRM analysts construct statistical/data mining models to better understand, as well as predict customers' future behavior. This makes customer intelligence very actionable, because we are able to quantify, e.g. the probability a customer is going to stop purchasing a firm's products/services in the coming year. This 'propensity to churn or attrite' can then be used to rank their entire customer base in order to prioritize which customers should receive special customer retention treatment. Using targeted marketing campaigns, analytical CRM empowers companies to learn from their campaigns, and finetune offers to different customer segments. Hence, businesses often turn to customer intelligence to increase their marketing ROI (return on investment). This enables marketing departments to become more accountable.

Many studies have shown that a good way to improve customer retention is to sell more products to the same customers, i.e. extend the portfolio of products or services bought from a given supplier. Of course, given a specific customer profile, companies would like to know what is the next most-likely product or service a customer is going to buy. In analytical CRM, we build cross-sell/up-sell models, also known as NPTB (next product to buy) models. These enable marketing analysts to target customers with the most appropriate product.

Customer intelligence goes beyond mere 'business intelligence', which is interpreted by software vendors as just report generation, and/or OLAP applications used to find reasons for deviations or above/below-average performance of businesses. While this approach is valuable, customer intelligence goes beyond reporting about the past. It empowers analysts to predict the most likely future events of individual customers.

In the Master of Science in Data Science for Business, we teach the analytics to build these analytical CRM models for retention, cross-sell, marketing optimization, ... . This gives our graduates a real advantage on the job market, because these are sought-after skills in today's competitive markets.

Academic papers as well as videos about customer intelligence can be found at www.crm.UGent.be.