Evaluation
Our project is finished. However, when we train a machine learning algorithm like OCR, for example, we need to know the best features and parameters to use and how to correct the classification, recognition, and detection errors in our system.
We need to evaluate our system with different situations and parameters and evaluate the errors produced in order to get the best parameters that minimize those errors.
In this chapter, we evaluated the OCR task with variables: size of low-level resolution image feature and the number of hidden neurons in the hidden layer.
We created the evalOCR.cpp application where we uses the XML training data file generated by the trainOCR.cpp application. The OCR.xml file contains the training data matrix ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access