Business Intelligence for New-Generation Managers

Business Intelligence for New-Generation Managers

1. Chapter 2: On the Way from a Knowledge Discovery in Databases to a Predictive Analytics

2. 1. Introduction

   • it is the task of decision makers to specify goals, to evaluate alternatives, and to choose between possible alternatives
     • This is supported by Business Intelligence (BI) systems, which are described as abilities, technologies, applications, and procedures for supporting enterprise activities to assist the understanding of the business environment

3. 2. Practical Implications of a Business Intelligence Based Forecast Process

   • 2.1 Forecasts within the Decision Process
     • 1. Provision of information (information acquisition and problem classification). 2. Design (determination of relevant criteria of the alternative selection). 3. Selection (determination of the most effective alternative).
   • 2.2 Business Intelligence in the Forecast Process
     • The determination, selection, and usage of relevant information and presented results depend on the respective user
   • 2.3 Problem Statement
     • In particular practice oriented journals discuss regularly that the quality of the used forecast models is not sufficient.

4. 3. Conceptual Modification of the KDD Process to Improve the Forecast Quality

   • 3.1 Knowledge Discovery in Databases Process Model
   • 3.2 KDD Enhancements in Context of a Predictive Analytics
     • An improvement in relation to the values of the classical KDD is obtained in particular by the extension and correction of the forecast model around future oriented data. In order to illustrate this correction calculation in context of an extended KDD processing concept, different possibilities exist.

5. 4. Case Study: Customer Relationship Management in a German Saving and Loans

   • The case is based on anonymous customer data of a German savings and loans bank
     • Appropriate products are offered according to the specific interests and points in time to be able to manage the relationship to the customer during his/her lifetime (in the sense of the CLV). I

6. 5. Conclusions

   • An implemented forecast model offers a systematic support, which is essential for successful enterprises

7. Me pareció una lectura interesante y toca un tema que no había pensado y el uso de árboles de decisiones, los cuales aprendimos que se podían utilizar para problemas simples y una vez más los vemos aplicados para solucionar cuestionamientos más avanzados.