Intelligent Predictive Analysis for Commodities (IPAC)

The Intelligent Predictive Analytics for Commodities (IPAC) Project will demonstrate a firstto-
market prediction-based data analytics capability for the managing of procurement
purchases with insight on the source prices of commodities and raw materials. This will be
achieved by applying novel data mining techniques to an extensive archive of commodities
prices and contextual data to establish trend characteristics. A key capability will be the price
prediction for combinations of commodities and raw materials to identify the lowest,
combined price.
The innovation in the IPAC Project is based on unique:
a) Trend-based data analytics to provide accurate and timely predictions on the expected
variations of the prices of commodities and raw materials;
b) ‘What If …’ analysis capability from the aggregation of information from a significant data
archive with live information on prices combined with a probabilistic based prediction
capability;
c) Visualisation algorithms that process the output from the predictive trend algorithms and
render the information in forms that can be easily manipulated by a user to enable useful
‘What if…’ analyses.
A recent state-of-the-art product evaluation has shown there is no other commercial solution
providing a predictive-based pricing of commodities and raw materials. The benefits for users
therefore are:
a) Improved understanding of commodities and raw materials prices enabling subscribers to
significantly decrease their procurements costs;
b) Subscribers will be able to undertake more sophisticated price analysis strategies by
identifying the consequences of combinations of aggregated commodity and raw materials
combinations. This will enable better planning of procurement activities.
The key outcome will be a prototype demonstrator produced and evaluated to confirm the
approach and to ensure the required predictive functionality and performance is achieved.
Economic benefit to both subscribers and consumers should follow.