Return to Science Search homepage
Return to Project List
|
Development of a Novel Belief Rule-based Methodology for Measuring and Assessing Food and Drink Quality - AFM 222 - FT1520
|
Description
The project aims at developing a novel belief rule-based methodology to support quality analysis and consumer acceptance prediction in rapid retro-design and testing of new food and drink products. |
Objective
To investigate existing decision support methods developed elsewhere for food product design and manufacture in order to evaluate and benchmark the proposed novel belief rule-based methodology.
To develop an initial consumer preference model for prioritising and synthesising consumer preferences using the evidential reasoning approach.
To use data mining methods and subjective expert knowledge to build an exploratory basic belief rule base representing essential causal relationships between consumer demands and product characteristics.
To research an optimal learning method to train and update the belief rule base into an optimal functional model using input-output data.
To research an adaptive optimisation method for setting the target values of product characteristics by optimising overall consumer preferences based on the models developed above.
To evaluate the above belief rule-based methodology for food products selected by the collaborating companies.
To benchmark the proposed methodology against the current methods such as artificial neural networks and linear regression and identify its desirable features particularly in terms of prediction accuracy and time required for training.
To write a full proposal for a Food LINK project to enable the commercial development and exploitation of this novel methodology based on the findings from this LINK Feasibility Study project.
|
Time-Scale and Cost
From:
2006
|
To:
2006
|
Cost: £49,993 |
|
Contractor / Funded Organisations
United Biscuits, Intelligent Decision Systems Ltd, Duckworth Flavours, Leatherhead Food International Ltd, University - Manchester |
Keywords
|
Fields of Study
Resource Efficient and Resilient Food Chain |