How to increase number of iterations for SVM Classifier

PC: HPZ440
Ram: 32GB
Processor: Intel(R). Xeon(R) CPU E5-1630 v3 4 Cores 8 Logical Processors

QGIS version: 3.22.7
OTB version: 8.1.1

So, when I run the train vector classifier using SVM as my chosen classification method I run out of iterations for running my process. In all the Orfeo documentation I can find and also YouTube videos I cannot seem to find a way to increase the number of iterations available. Is there any way of achieving the number of iterations so all my training data is utilized?

Thanks in advance
Lucas

Hi @lmcmaster

Thanks for reporting the problem, indeed I think it is not possible to increase the number of iterations, it could be a nice feature to implement in future OTB versions. You can open an issue in the orfeo-toolbox gitlab

Best regards

It seems you’re facing an issue with running the train vector classifier using SVM in QGIS with Orfeo Toolbox (OTB). The number of iterations for the training process might be limited, and you’re looking for a way to increase this limit.

Unfortunately, the issue you’re encountering might be specific to the implementation details of the SVM classifier in the software or the version you’re using. To address this:

  1. Check QGIS and OTB Documentation: Verify if there’s any information in the official documentation of QGIS and Orfeo Toolbox regarding SVM training parameters and iteration limits. Sometimes, this information can be buried in the documentation.
  2. QGIS and OTB Forums: Visit the QGIS and OTB forums or community pages. Users and developers often discuss such issues, and you might find relevant threads or be able to post your question for assistance.
  3. Software Updates: Ensure that you’re using the latest versions of QGIS and OTB. Updates may include bug fixes or improvements to the SVM classifier implementation.
  4. Contact Developers: If the issue persists, consider reaching out to the developers or maintainers of QGIS and OTB. They might be able to provide insights or address the problem in future updates.

Unfortunately, without specific details on SVM training parameters and iteration limits in the QGIS and OTB versions you’re using, I can’t provide a precise solution. Consulting the documentation and community forums would be a good starting point.