DAPR in Finance:
Successful tests have been conducted in the following areas:

• Best execution compliance automation

• Liquidity prediction

• Non-STP correction

• Customer targeting

• Operational risk and losses

• Trade performance benchmarking

PropheZy consists of the following components:

• an Excel add-in comprising the user interface, help files, data manipulation and reporting;

• a classifier and modeller for building a predictive engine based on probability density estimation;

• a server-side predictive engine manager capable of handling hundreds of simultaneous users requesting predictions.

PropheZy is based upon sophisticated statistical techniques that classify and model multivariate data to make predictions on new data.  PropheZy works best on continuous numerical data but can handle discrete, categorical data and time series data.  PropheZy improves on traditional approaches such as neural networks and provides information about the classification confidence levels, so that appropriate human choices can be made.

PropheZy can be rolled-out to thousands through a browser interface by running an ASP service.  Applications can be updated frequently to take account of changing conditions.  PropheZy is inherently parallel, thus ensuring that successful applications are scalable for complex, real-time applications.  Potential applications include increasing success rates among targets; reducing customer churn; identifying best sales techniques, bid approaches or cross-sales; setting targets for client revenue, profitability, client satisfaction. 

PropheZy has been extensively tested using standardised datasets - its performance is amongst the best available.

PropheZy is often used in conjunction with VizZy – a visualisation package which allows data to be more easily analysed and understood.

More information:

PropheZy flyer

PropheZy overview presentation

PropheZy - a television case study

PropheZy articles

Michael Mainelli, "Predictive Compliance in Financial Services", New Lens, Fusion Experience (November 2009).

Michael Mainelli and Mark Yeandle, "Best Execution Compliance: Towards An Equities Compliance Workstation , Journal of Risk Finance, Volume 7, Number 3, pages 313-336, Emerald Group Publishing Limited (June 2006).

Michael Mainelli and Mark Yeandle, "Best Execution Compliance: New Techniques for Managing Compliance Risk" , Journal of Risk Finance, Volume 7, Number 3, pages 301-312, Emerald Group Publishing Limited (June 2006).

Mark Yeandle, "Best Execution Compliance Automation" , Financial Services Research pages 80-82, Financial Services Research Limited (April 2006).

Ian Harris, Michael Mainelli, Peter Grant and Jenny Harrow, "Predicting The Effectiveness Of Grant-Making", Journal of Strategic Change, Volume 15, Number 2, pages 53-66, John Wiley & Sons (March-April 2006).

Michael Mainelli, "Competitive Compliance: Manage and Automate, or Die",Journal of Risk Finance, The Michael Mainelli Column, Volume 6, Number 3, pages 280-284, Emerald Group Publishing Limited (June 2005).

Michael Mainelli, “Benchmarking Facilities Management”, Essential FM Report, Number 43, pages 6-7, Tottel Publishing (January 2005).

Michael Mainelli, “Finance Looking Fine, Looking DAPR: The Importance of Dynamic Anomaly and Pattern Response”, Balance Sheet, The Michael Mainelli Column, Volume 12, Number 5, pages 56-59, Emerald Group Publishing Limited (October 2004).

Z/Yen Seizes British Computer Society Award for PropheZy prediction engine(October 2004)

DTI Bestows Smart Award on Z/Yen for PropheZy Prediction Engine

Jean Haggerty, "Banks Test Operational Risk Tools", International Financing Review (26 July 2003). 

Michael Mainelli, Ian Harris and Alan Helmore-Simpson, "The Auditor's Cross Subsidy" (statistical modelling of audit prices), Strategic Planning Society E-Newsletter, Article 1 (June 2003).

VizZy