Project Details
Description
Why is this interesting scientifically?
The developed methodology:
• provides a novel generic framework that enables quick evaluation
and remediation of the forecasts and/or simulations based on
parametric models or domain experts.
• closes the knowledge gap of scientifically integrating the parametric
statistical models with non-parametric machine learning models
through model fusion.
• is a semi-parametric tool (consisting of both parametric and nonparametric
components), it, therefore, contributes to the stream of
explainable-artificial-intelligence literature
How is this relevant to the materials transition?
One what issues would you like to get input from others?
Can improvements be found in the way:
• our models captures new information?
• our experts absorb new information?
• our data management systems distribute new information?
The developed methodology:
• provides a scientifically valid and practically efficient tool to engage
stakeholders and experts in the dialogue of projecting the
future transition scenarios for textile and building materials
sectors.
• can be formulated as a decision-support tool that combines the
knowledge of different experts, and parametric, and non-parametric
approaches to facilitate informed policy decisions regarding the
bioeconomy transition.
• will be used as course material in a new MSc course provided by
WU-AEP on machine-learning time-series forecasting.
What are the key activities or steps?
• Developing a methodology that can be used to determine to what
extent a specific approach (e.g. expert opinions, parametric models,
etc.) is fast and flexible enough to adopt new information.
• Organising a workshop for bioeconomy experts in the textiles
and building sectors to discuss the developed methodology
• Upgrade the methodology based on the workshop
• Apply the methodology to the selected empirical case of the
workshop
• Write a manuscript about the results of the empirical case (we
write the manuscript parallel to the empirical analysis)
Status | Finished |
---|---|
Effective start/end date | 1/01/23 → 31/12/23 |