MBAS903

Purpose: Preparing a report for your client to gain insights into the characteristics

of different types of houses sold.

Business Analytics Applications: descriptive analytics, cluster analysis

Background: The AMESHOUSING3 dataset contains data on the characteristics of 300 houses sold. The client would like to understand the different grouping of houses sold.

Guidelines: Use only housing characteristics data (ie. no sales price variables) for the cluster analysis. When choosing variables you need to use results from your descriptive analytics of your variables (central tendency, variation, shape, correlations). Develop a range of cluster analyses and experiment with differing numbers of clusters and variables. As a final step, analyse the relationship between your clusters and "Sale price in dollars". DO NOT include sales price in the cluster analysis itself . There is no one "right' answer, so think about and discuss in your report the rationale for the steps you take in the analysis and the reasoning for your preferred cluster model.

Format: Business report for client

Executive summary - one paragraph overview of full project including summary of results - 10 marks

Introduction - set the scene and background of client issue/problem (broad to specific issues) - 10 marks

Methodology - discuss the rationale and planned steps for your analytics process (ie. details of descriptive analytics and cluster analysis you plan to do - DO NOT include any results in this section) - 20 marks

Body and analysis - descriptive analytics results, one paragraph summary of the process and steps ultimately taken in your cluster analysis (variables and k experimented with), on paragraph summary justifying your choice of final model, then present and interpret model results (DO NOT waste space presenting other models that are not your chosen one), examine possible link with sale price variable briefly. - 50 marks

Conclusions - wrap up the exercise with recommendations for client -