Machine Learning and Visual Analytics for Consulting Business Decision Support

The application of machine learning and statistical predictive models to business problems has found success in recent years due to an exponential increase in consumer data and readily available computational power. However, visualising and interpreting this data to support business decision making in the context of consulting businesses is challenging and there is scope for advancement. The accurate prediction of hours to be spent on a project (cost) ahead of time underpins the profitability of these organisations. The aim of the research is twofold: to identify suitable techniques from the fields of machine learning and statistics for internal cost prediction in a consulting business, and to develop a user interface with visual analytics displaying results from these techniques to provide interactive decision support. The data for this project was collected from a consulting business' customer relationship management (CRM) database, which contained twelve years of past consulting projects. To date, statistical linear models and machine learning decision trees have been trialed and the research will progress into random forests, neural networks, and support vector machine (SVM) models. A prototype user interface and visualisation of the results has also been developed.