AI Skill Classification

RE: AI Skill Classification

SRINIVAS M — Thu, 08 Feb 2024 13:00:03 GMT

To address the issue of low confidence scores for untrained document types, it's essential to:

Clearly separate known and unknown document types during the training phase.
Consider training the model to recognize an "other" category for documents that do not fit into the predefined types.

RE: AI Skill Classification

Melvin George — Thu, 08 Feb 2024 12:47:03 GMT

Thanks Stefan. i upload a picture in the PDF which is possible scenario in our process. we decide the flow based on the confident score but now it is different .. will check further

RE: AI Skill Classification

Abhay Dalsaniya — Tue, 06 Feb 2024 19:18:28 GMT

May be Appian is still able to identify some of the fields from the non-trained document.

RE: AI Skill Classification

Stefan Helzle — Tue, 06 Feb 2024 10:05:39 GMT

I think this happens when the identified patterns in the documents are not strong enough and you upload a document that does not deviate enough from the training material.

RE: AI Skill Classification

Melvin George — Tue, 06 Feb 2024 09:19:23 GMT

Thanks Abhay but my curiosity is Why confident score is still 50 when non trained document uploaded? confident score is crucial to decided on the flow of the process.

RE: AI Skill Classification

Abhay Dalsaniya — Mon, 05 Feb 2024 19:33:32 GMT

The confidence score is internal to Appian and it represents how close the document classification and extraction is.

RE: AI Skill Classification

Melvin George — Mon, 05 Feb 2024 10:36:39 GMT

Thank you gabby. I will check and revert if i have questions

RE: AI Skill Classification

gabby.rothschild — Thu, 01 Feb 2024 21:59:26 GMT

Hi Melvin,

Because machine learning is not deterministic, a high variance in your confidence score or accuracy scores may mean that your dataset is not representative enough. Please refer to our documentation on how to create a representative dataset: https://docs.appian.com/suite/help/23.4/collect-data.html as well as how to evaluate model performance: https://docs.appian.com/suite/help/23.4/evaluate-ai.html