Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Do you really want to delete this prezi?
Neither you, nor the coeditors you shared it with will be able to recover it again.
Make your likes visible on Facebook?
You can change this under Settings & Account at any time.
Copia de Explain Any Topic
Transcript of Copia de Explain Any Topic
FEED FEED INGREDIENTS
: create an engine which can predict, from a spectra file, which group this spectra belongs to.
-First, we need a classification for each group.
-Based on the classification, models are built for each step when removing groups.
-The engine is built placing these models one after another; they act like filters when a spectra file is entered.
-Engine works in a hierarchical (step-wise) manner.
-Depending on the characteristics of each spectra, the models will guide the files to a final solution, which will be then displayed.
An NIRS Prediction Engine for Discrimination of Animal Feed Ingredients
Calibration / Validation samples
100 representative samples were taken for calibration and the rest for validation in each of the groups. Kennard-Stone algorithm was used, based on spectral data.
Reason: We want to have a similar number of samples of all the groups and we want them to be representative of their whole group.
Aitziber Miguel Oyarbide
Universidad Autónoma de Madrid
Who is AUNIR?
The leading company in NIR calibration
development and delivery.
AUNIR provides NIR solutions and calibration requirements for the animal feed industry all over the world.
-Pre-processing method that is applied to the spectra before performing PCA.
-Covariance filter: identifies patterns in the variables which should be removed.
-Down-weighing of features that make the data differ.
-Mean centers each class, removes between-class variations and orthogonalises to the clutter subspace.
-It reduces the number of factors required for discriminant modelling.
High Protein High Oil
High Protein Low Oil
Low Protein High Oil
Low Protein Low Oil
-Classification and building of a prediction engine for more groups of our library.
-Improvement of the accuracy of prediction engines.
-Expansion of the application of this project to all groups in our database.