Module 1&2 Workshops and Assignment Copyright © Jacob L. Cybulski M2&3 Assignment Preview: SAS EM, Regression, NN, DTs The following mini-case study will be used in your group assignment A2. Groups are up to 3 members (1-2-3). An independent online business Best Iowa Buys have setup a members-only service to predict the likely value of real-estate put up for auction. They have collected sample data about property sales in Ames, Iowa (USA), in CSV format. The data consists of 2,930 records of properties sold from 2006 till late 2010 in Ames, each described with 79 variables. You have been asked to develop an analytics solution to estimate the price of Ames properties advertised for auction. They are also interested in the property classification in terms of its affordability within its category or group, and its value for money for the potential buyer. Select the best predictive model and provide a summary of the model and its performance. 1 Software: Access SAS EMiner on Deakin Apps On Demand prior to classes Explore: Learn SAS EMiner, explore the data, tabulate and chart your findings Cluster Analysis: Perform cluster analysis and segmentation of your data to identify categories or groups of the Ames properties that could be used to guide the customer buying choices Modelling and Evaluation: Create a number of predictive models (e.g. Neural Nets, Regression and Decision Trees) based on the available data, evaluate, possibly using cross-validation techniques, and optimise their performance Model Integration: Create an analytic solution, possibly involving ensemble models, integrating and evaluating the models developed so far Submission: Present, explain and justify all results and recommendations, report your final results and final findings as a team via CloudDeakin