I'm working on developing an disease/early warning system for cocoa disease to provide timely forecasts that farmers can use to mitigate potentially damaging events through preventative management. Therefore, I would like to get information on methods available besides regression to develop the model and also verify the developed model.
Hi Albert,
There is no simple answer to your question. As indicated in Alessandro's answer, there are many potentially useful methods, the most suitable one will depend on many factors including (but not limited to):
The type of disease you want to predict
The availability of different types of information
The availability of information technology to develop and support a model
The forecast interval (i.e. how far in advance you want to make the predictions)
The capacity of growers to make use of different types of information
The budget available for developing and testing the system
The level of accuracy required
The costs of operating the system
The costs of mistaken predictions (both false positives and false negatives)
Yours,
N
Hi Alessandro and Neil,
Thanks a lot for the information provided. I did agree with Neil as the scope for developing the disease warning system need to consider a lot of factors. Currently, I'm working on developing the disease prediction model on cocoa black pod disease caused by the Phytopthtora Palmivora sp. The X factors to be considered for the prediction will be the vapour pressure deficit (VPD), inoculum level and also the pod development stage and the Y will be the disease incidence.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics