Business Analytics using Data Mining

Description

Overview: Modelling business data to mine for information is a skill in demand by our graduate employers. BA graduates are expected to understand the basic principles and theory of the methods used to model and manage data as well as being able to apply the methods to business problems. This class seeks to develop both the knowledge and skills to achieve these goals. The class seeks to provide students with the opportunity to develop analytical approaches for mining data using commercial software that will be intellectually challenging and useful. This class focuses on the methods and tools used for mining data, drawing upon a range of business problems. Syllabus: This class will begin by introducing data mining problems, models and solutions to set the scene. The principles of statistical and rule based methods will be introduced, particularly for clustering, classification and predictive problems, and applications to case studies will be illustrated using commercial software packages. Case studies will include, for example, classification of telecom company customers, development of credit scoring rule for financial organisation and development of knowledge base system. External speakers from industry and the software providers will present insights into the large scale business applications and the developments within the field. Learning Outcomes: Subject specific knowledge and skills:  To understand the purpose and use of data mining in business;  To understand the basic principles and theory of statistical and artificial intelligence methods used in data mining;  To develop appropriate methods and models for mining different types of business data and problems;  To use appropriate commercial software tools to implement data mining. Cognitive abilities and non-subject specific skills:  To develop personal communication skills, both verbal and written  To work in teams  To enhance analytical skills Assessment: Deadlines Group = 23 April 2012 Individual = 8 May 2012 10% of marks will be allocated for group case study presentations that will be aligned with specific problems of clustering, classification and prediction introduced in class and partly supported by tutored labs. The aim is to provide immediate feedback on the understanding and implementation of methods for relatively well bounded problems. 40% of marks will be given for a group assignment to develop and instantiate a knowledge base system to support decision-making. Students will be required to submit their knowledge base as well as supporting documentation. 50% of marks will be given for an individual assignment that will involve analyzing a business problem using appropriate methods implemented using statistical software and preparing a suitable report of the analysis that answers the questions posed by the client. The individual assignment will also involve a reflective element to compare and contrast different classes of data mining methods for the selected problem.